Security Manager Guide Video Surveillance v2 0

Security Managers Guide to Video Surveillance
Version 2.0 / November 2008 John Honovich IPVideoMarket.Info
Contents
I. Introduction/Tutorials on Video Surveillance Chapter 1: How to Design Video Surveillance Solutions...................................................................8 Chapter 2: Introduction to V!s / I" Video So#tware.....................................................................1$ Chapter %: Introduction to Video C&D'Cs......................................................................................2( Chapter (: )andwidth )asics #or Video Surveillance.......................................................................2$ Chapter *: '+a,ining Video -nal.tics.............................................................................................%( Chapter /: 0icense "late !ecognition Tutorial.................................................................................%8 Chapter 1: Introduction to DV!/ V! Storage &pti,i2ation..........................................................(2 Chapter 8: 3ireless Video Surveillance Tutorial..............................................................................(8 Chapter $: -"I and S.ste, Integration Tutorial...............................................................................*% Chapter 14: How to Integrate Video 3ith &ther S.ste,s................................................................*1 Chapter 11: How to 5igrate #ro, -nalog to I" Ca,eras................................................................/4 Chapter 12: Director. o# &n60ine Video Surveillance Tutorials....................................................../* II. 'valuating 7e. Trends and Technologies Chapter 1%: IT Is ot Ta8ing &ver Securit.....................................................................................11 Chapter 1(: 3ill Securit. Integrators Survive9................................................................................1( Chapter 1*: Should I :se I" Ca,eras9............................................................................................1$ Chapter 1/: Top * "ro;le,s o# I" Ca,eras.....................................................................................8( Chapter 11: Value o# H.;rid DV!s/ V!s.......................................................................................81 Chapter 18: ew &ptions #or DV!/ V! Storage...........................................................................$4 Chapter 1$: The Value o# Centrali2ed Video -nal.tics....................................................................$1 Chapter 24: '+a,ining <&pen< S.ste,s...........................................................................................$$ Chapter 21: The Danger o# )u.ing "ac8ages.................................................................................142 Chapter 22: Introduction to Cit.63ide Surveillance......................................................................14( Chapter 2%: Is "u;lic CCTV '##ective9.........................................................................................148 III. 'valuating ew "roducts Chapter 2(: How to !ead 5ar8eting 5aterial................................................................................118 Chapter 2*: How to 'valuate ew Technolog...............................................................................121 Chapter 2/: How to Calculate Video Surveillance !&Is................................................................128
About IP Video Market Info

Personalized News
Want to stay informed of new trends, analysis and technology for video surveillance? IP Video Market Info offers the most extensive daily and deep coverage in the industry. top !y the site or sign up for personali"ed news.
Expert Help
Making a decision on new systems or evaluating changes or upgrades? IP Video Market runs a marketplace of the world#s leading experts availa!le to help you $uickly, simply and inexpensively make the !est decision.
2009 Industry Guide

If you want more in%depth competitive and industry analysis, consider our &''( Video urveillance Industry )uide. *his )uide provides recommendations on how to plan and respond to the recession, evaluating what technologies, products and companies will do the !est in &''(. *he &+, page guide can !e downloaded and printed out for easy reading. - premium product, the cost is as low as .+'.'' / 0 for personal use and .&''.'' / 0 to share with anyone in your company.
About the Author

1ohn 2onovich is the founder of IP Video Market Info, the leading we!site dedicated to video surveillance. 1ohn researches and writes extensively for IP Video Market Info, providing ongoing and timely analysis of new technologies and emerging products. -dditionally, 1ohn developed software that allows IP Video Market Info to constantly track and organi"e new video surveillance information from company we!sites and across the we!. Prior to founding IP Video Market Info, 1ohn was a successful manager and engineer working closely with ecurity Managers to develop video surveillance solutions. -s 0irector of Product Management for 3V4 ecurity, 1ohn helped design and deploy industry leading video analytic and facial recognition software for the !anking and retail market. -s )eneral Manager of ensormatic 2awaii, 1ohn lead large scale military and critical infrastructure deployments of video analytics, IP video and wireless video surveillance. 5efore entering the Physical ecurity industry, 1ohn was a senior engineer designing IP Video over 0 6 networks for telecommunication carriers. 1ohn graduated from 0artmouth 7ollege and, over the years, has achieved 7isco certifications and the - I International 5oard 7ertification in Physical ecurity 8P P9.
Preface
Who is this Book for? *his !ook is designed for the security manager who uses video surveillance:77*V systems. ;ou should !e a!le to understand this !ook if you have used a 0V4 system. *he !ook#s goal is to help you make !etter decisions a!out evaluating and selecting video surveillance systems. Integrators and manufacturers should also !e a!le to learn from this, especially to gain a !etter appreciation of drivers for security managers. Can I Share this Book with Others? ;es. *his is a free and <open source= !ook. ;ou can share and copy the !ook as long as you attri!ute the source 81ohn 2onovich, IPVideoMarket.info9 and do not restrict other#s a!ility to share the !ook. *his is technically called a <7reative 7ommons -ttri!ution% hare -like 3.' /nported 6icense.= >mail me at ?honovich@ipvideomarket.info with any $uestions. Will this Book be Updated? ;es, this !ook will !e updated & to 3 times per year and is designed to !e a living !ook that reflects ongoing developments in video surveillance. )o to httpA::ipvideomarket.info:!ook to check for updates. Can I Suggest Improvements or New Topi s for the Book? ;es, I strongly encourage you to suggest improvements or new topics. Please email me at ?honovich@ipvideomarket.info.
I Introduction to Video Surveillance
How to Design Video Surveillance Solutions
Chapter 1:
Designing a video surveillance solution requires decisions on 7 fundamental questions. This tutorial walks the reader through each issue explaining the basic options and the rationale for selecting different options. This is a survey to help those new to video surveillance. Its goal is to quickly identify the key aspects of video surveillance design, not to examine the many details and edge cases in such designs. The 7 fundamental questions are

!hat type of cameras should I use" #ow should I connect cameras to video management systems" !hat type of video management system should I use" !hat type of storage should I use" !hat type of video analytics should I use" #ow should I view my surveillance video" #ow should I integrate video with my other systems"
1. Cameras $ameras are literally the eyes of a video surveillance system. $ameras should be deployed in critical areas to capture relevant video. The two basic principles of camera deployment are %&' use chokepoints and %(' cover assets.
I"Video5ar8et.In#o 1
$hokepoints are areas where people or vehicles must pass to enter a certain area. )xamples include doorways, hallways and driveways. *lacing cameras at chokepoints is a very cost+effective way to document who entered a facility. ,ssets are the specific ob-ects or areas that need security. )xamples of assets include physical ob-ects such as safes and merchandise areas as well as areas where important activity occurs such as cash registers, parking spots or lobbies. !hat is defined as an asset is relative to the needs and priorities of your organi.ation. /nce you determine what areas you want to cover, there are 0 camera characteristics to decide on &. Fixed vs PTZ , camera can be fixed to only look at one specific view or it can be movable through the use of panning, tilting and .ooming %i.e., moving left and right, up and down, closer and farer away'. 1ost cameras used in surveillance are fixed. *T2 cameras are generally used to cover wider fields of views and should generally only be used if you expect a monitor to actively use the cameras on a daily basis. , key reason fixed cameras are generally used is that they cost 3 +4 times less than *T2s %fixed cameras average 5(66 + 5366 78D whereas *T2 cameras can be over 5(,666 78D'. (. Color vs Infrared vs Thermal: In T9, a video can be color or black and white. In video surveillance today, the only time producing a black and white image makes sense is when lighting is very low %e.g., night time'. In those conditions, black and white images are produced by
I"Video5ar8et.In#o
infrared or thermal cameras. Infrared cameras require special lamps %infrared illuminators' are fairly inexpensive for producing clear image in the dark. Thermal cameras require no lighting but product only outlines of ob-ects and are very expensive %53,666 + 5(6,666 on average' In day time or lighted areas, color cameras are the obvious choice as the premium for color over black and white is trivial. %. Standard Definition vs. Megapixel: This choice is similar to that of T9s. :ust like in the consumer world, historically everyone used standard definition cameras but now users are shifting into high definition cameras. !hile high definition T9 maxes out at ; 1*, surveillance cameras can provide up to &< 1* resolution. In (664, megapixel cameras only represent about 0= of total cameras sold but they are expanding very rapidly. 8ee a demonstration of megapixel cameras to learn more. 0. IP vs Analog: The largest trend in video surveillance today is the move from analog cameras to I* cameras. !hile all surveillance cameras are digiti.ed to view and record on computers, only I* cameras digiti.e the video inside the camera. !hile most infrared and thermal cameras are still only available as analog cameras, you can only use megapixel resolution in I* cameras. $urrently, (6= of cameras sold are I* and this percentage is increasingly rapidly. 1ost organi.ations will mix and match a number of different camera types. >or instance, an organi.ation may use infrared fixed analog cameras around a perimeter with an analog *T2 overlooking the parking lot. /n the inside, they may have a fixed megapixel camera covering the warehouse and a number of fixed I* cameras covering the entrance and hallways.
I"Video5ar8et.In#o
. Conne!tivit" In professional video surveillance, cameras are almost always connected to video management systems for the purpose of recording and managing access to video. There are two main characteristics to decide on for connectivity.
IP vs. Analog 9ideo can be transmitted over your computer network %I*' or it can be sent as native analog video. Today, most video feeds are sent using analog but migration to I* transmission is rapidly occurring. ?oth I* cameras and analog cameras can be transmitted over I*. I* cameras can connect directly to an I* network %-ust like your *$'. ,nalog cameras cannot directly connect to an I* network. #owever, you can install an encoder to transmit analog feeds over I*. The encoder has an input for an analog camera video feed and outputs a digital stream for transmission over an I* network. @earn more about the choice between I* and analog transmission. #ired vs #ireless 9ideo can be sent over cables or though the air, whether you are using I* or analog video. /ver A6= of video is sent over cables as this is generally the cheapest and most reliable way of sending video. #owever, wireless is an important option for transmitting video as deploying wires can be cost+prohibitive for certain applications such as parking lots, fence lines, remote buildings. @earn more about when and how to use wireless video surveillance.
$. %ideo Management S"stem
I"Video5ar8et.In#o
14
9ideo management systems are the hub of video surveillance solutions, accepting video from cameras, storing the video and managing distribution of video to viewers. There are 0 fundamental options in video management systems. 1ost organi.ations choose & of the 0. #owever, as companies may have multiple types when they transition between one and another.
D%&s are purpose built computers that combine software, hardware and video storage all in one. ?y definition, they only accept analog camera feeds. ,lmost all D9Bs today support remote viewing over the Internet. D9Bs are very simple to install but they significantly limit your flexibility in expansion and hardware changes. D9Bs are still today the most common option amongst professional buyers. #owever, D9Bs have definitely fallen out of favor and the trend is to move to one of the ; categories below. 'D%&s or hybrid D9Bs are D9Bs that support I* cameras. They have all the functionality of a D9B listed above plus they add support for I* and megapixel cameras. 1ost D9Bs can be software upgraded to become #D9Bs. 8uch upgrades are certainly a significant trend and is attractive because of the low migration cost %supports analog and I* cameras directly'. @earn more about the value and issues in selecting HDV!s. (%&s are like D9Bs in all ways except for camera support. !hereas a D9B only supports analog cameras, an C9B only supports I* cameras. To support analog cameras with an C9B, an encoder must be used. IP %ideo S)rveillan!e Soft*are is a software application, like !ord or )xcel. 7nlike D9Bs or C9Bs, I* 9ideo 8urveillance 8oftware does not
I"Video5ar8et.In#o
11
come with any hardware or storage. The user must load and set up the *$D8erver for the software. This provides much greater freedom and potentially lower cost than using D9BDC9B appliances. #owever, it comes with significant more complexity and time to set up and optimi.e the system. I* 9ideo 8urveillance 8oftware is the hottest trend in video management systems currently and is the most frequent choice for very large camera counts %hundreds or more'. @earn more about choosing so#tware onl. s.ste,s. +. Storage 8urveillance video is almost always stored for later retrieval and review. The average storage duration is between ;6 and A6 days. #owever, a small percentage of organi.ation store video for much shorter %7 days' or for much longer %some for a few years'. The two most important drivers for determining storage duration is the cost of storage and the security threats an organi.ation faces. !hile storage is always getting cheaper, video surveillance demands huge amount of storage. >or comparison, EoogleFs email service offer about 7 E?D s of free email storage. This is considered to be an enormous amount for email. #owever, a single camera could consume that much storage in a day. It is fairly common for video surveillance systems to require multiple T?s of storage even with only a few do.en cameras. ?ecause storage is such a significant cost, numerous techniques exist to optimi.e the use of storage.
I"Video5ar8et.In#o
12
The type of security threats also impact determining storage duration. >or instance, a ma-or threat at banks is the report of fraudulent investigations. These incidents are often not reported by affected customers until <6 or A6 days after the incident. ,s such, banks have great need for longer term storage. ?y contrast, casinos usually know about issues right away and if a problem is to arise they learn about it in the same week. $asinos then, very frequently, use much shorter storage duration %a few weeks is common'. Three fundamental types of storage may be selected &. Internal storage is the hard drives that are built inside of a D9B, C9B or server. This today is still the most common form of storage. !ith hard drives of up to & T? common today, internal storage can provide total storage of (T? to 0T?. Internal storage is the cheapest option but tends to be less reliable and scalable than the other options. Conetheless, it is used the most frequently in video surveillance. (. Dire!tl" Atta!hed storage is when hard drives are located outside of the D9B, C9B or server. 8torage appliances such as C,8 or 8,Cs are used to manage hard drives. This usually provides greater scalability, flexibility and redundancy. #owever, the cost per T? is usually more than internal storage. ,ttached storage is most often used in large camera count applications. %. Storage Cl)sters are I* based FpoolsF of storage speciali.ed in storing video from large numbers of cameras. 1ultiple D9Bs, C9Bs or servers can stream video to these storage clusters. They provide efficient, flexible and scalable storage for very large camera counts. 8torage clusters are the most important emerging trend in video surveillance storage. @earn more about storage clusters for video surveillance.
I"Video5ar8et.In#o 1%
,. %ideo Anal"ti!s 9ideo analytics scan incoming video feeds to %&' optimi.e storage or %(' to identify threateningDinteresting events. 8torage optimi.ation is the most commonly used application of video analytics. In its simplest form, video analytics examines video feeds to identify changes in motion. ?ased on the presence or absence of motion, the video management system can decide not to store video or store video at a lower frame rate or resolution. ?ecause surveillance video captures long periods of inactivity %like hallways and staircases, buildings when they are closed, etc.', using motion analytics can reduce storage consumption by <6= + 46= relative to continuously recording. 7sing video analytics to identify threateningDinteresting events is the more FexcitingF form of video analytics. Indeed, generally when industry people talk of video analytics, this is their intended reference. $ommon examples of this are perimeter violation, abandoned ob-ect, people counting and license plate recognition. The goal of these types of video analytics is to pro+actively identify security incidents and to stop them in progress %e.g., perimeter violation spots a thief -umping your fence so that you can stop them in real time, license plate recognition identifies a vehicle belonging to a wanted criminal so you can apprehend him'. These video analytics have been generally viewed as a disappointment. !hile many observers believe that video analytics will improve, the video
I"Video5ar8et.In#o
1(
analytics market is currently contracting %in response to its issues and the recession'. @earn more about the challenges o# video anal.tics.
-. %ie*ing %ideo 8urveillance video is ultimately viewed by human beings. 1ost surveillance video is never viewed. /f the video that is viewed, the most common use is for historical investigations. 8ome surveillance video is viewed live continuously, generally in retail %to spot shoplifters' and in public surveillance %to identify criminal threats. 1ost live video surveillance is done periodically in response to a Fcalled+inF threat or to check up on the status of a remote facility. 0 fundamental options exist for viewing video.
.o!al %ie*ing directly from the D9B, C9B or servers is ideal for monitoring small facilities on site. This lets the video management system double as a viewing station, saving you the cost of setting up or using a *$. This approach is most common in retailers, banks and small businesses. &emote PC %ie*ing is the most common way of viewing surveillance video. In this approach, standard *$s are used to view live and recorded video. )ither a proprietary application is installed on the *$ or a web browser is used. 1ost remote *$ viewing is done with an installed application as it provides the greatest functionality. #owever, as web applications mature, more providers are offering powerful web viewing. The advantage of watching surveillance video
I"Video5ar8et.In#o
1*
using a web browser is that you do not have to install nor worry about upgrading a client. Mo/ile %ie*ing allows security operators in the field to immediately check surveillance video. ,s responders and roving guards are common in security, mobile viewing has great potential. Though mobile clients have been available for at least 3 years, they have never become mainstream due to implementation challenges with *D,sDphones. Benewed interest and optimism has emerged with the introduction of the ,pple i*hone. @earn more about how -pple<s i"hone is i,pacting video surveillance. %ideo #all %ie*ing is ideal for large security operation centers that have hundreds or thousands of cameras under their -urisdiction. 9ideo walls provide very large screens so that a group of people can simultaneously watch. This is especially critical when dealing with emergencies. 9ideo walls generally have abilities to switch between feeds and to automatically display feeds from locations where alarms have been triggered.
0. Integrating %ideo *ith 1ther S"stems 1any organi.ations use surveillance video by itself, simply pulling up the video management systemsF client application to watch applications. #owever, for larger organi.ations and those with more significant security concerns, this is an inefficient and poor manner to perform security operations. Instead, these organi.ations prefer an approach similar to the militaryFs common operational picture %$/*' where numerous security systems all display on a singular interface. Three ways exist to deliver such integration with video surveillance
I"Video5ar8et.In#o 1/
A!!ess Control as ')/ 1ost organi.ations have electronicDI* access control systems. These systems have been designed for many years to integrate with other security systems such as intrusion detection and video surveillance. This is the most way to integrate video surveillance and relatively inexpensive %5&6,666 + 536,666 78D'. #owever, access control systems are often limited in the number and depth of integration they support. PSIM as ')/ In the last few years, manufacturers now provide speciali.ed applications whose sole purpose is to aggregate information from security systems %like video surveillance' and provide the most relevant information and optimal response policies. These applications tend to be far more expensive %%5&66,666 + 5&,666,666 78D' yet support a far wider range of security manufacturers and offer more sophisticated features. %ideo Managemenet S"stem as ')/ Increasingly, video management systems are adding in support for other security systems and security management features. If you only need limited integration, your existing video management system may provide an inexpensive %yet limited' solution. @earn more about options for integrating video with other systems.
Con!l)sion If you feel comfortable with the key decisions to be made, you may want to start examining what companies provide the best products for your need. Gou can learn more about companies for each component at the I* 9ideo 1arket $ompanies /verview directory.
I"Video5ar8et.In#o
11
Introduction to V!s / I" Video So#tware
Chapter 2:
Introduction to NVRs / IP Video So tware
I" Video Surveillance and etwor8 Video !ecorders = V!s> are two o# the ,ost co,,on ter,s descri;ing the use o# I" ca,eras and networ8 ;ased co,puters in ph.sical securit.. )oth o# these ter,s are ,ar8eting phrases and are not controlled ;. standards ;od.. -s such? no authoritative de#inition is possi;le and ,an. diverging opinions are held. This article atte,pts to docu,ent the ,ost agreed upon assu,ptions and highlight the ,ost widel. de;ated ele,ents. 5oreover? a de;ate e+ists in the industr. over what to call these solutions. !e#lecting the legac. o# DV!s? ,an. call these s.ste,s V!s. However? this ter, suggests hardware and proprietar. appliance. 5an. #eel strongl. that these solutions should ;e open architecture and <so#tware onl.<. -s such? ,an. do not consider their products to ;e < V!s<. @reAuentl. ,anu#acturers re#er to their products as BI" Video 5anage,entB solutions or BI" Video SurveillanceB solutions. @or purposes o# ;revit.? I use the acron., B V!B in this docu,ent instead o# the long and unwield. alternative ,anu#acturers suggest o# CI" Video surveillance so#tware? etcD. Do note that ,anu#acturers #eel ver. strongl. a;out the na,ing o# the categories their products are placed in. I would reco,,end .ou ignore the categor. na,es and #ocus on understanding the di##erences in ;ene#its. NVRs Must Support IP Cameras -l,ost ever.one agrees that to ;e designated an V! a solution ,ust support I" ca,eras. Indeed? the networ8 in Cnetwor8 video recorderD is generall. accepted as re#erring to the use o# an I" networ8 to connect I" ca,eras to an V!.
NVRs are Software Only Applications (DEBA ED! I"Video5ar8et.In#o 18
5ost V! suppliers o##er their product as so#tware onl.. That is to sa. the V! provides the user with #iles that are loaded on a co,puter o# the user<s choosing. The user does not have to purchase the hardware o# the V! supplier. This is widel. considered to ;e a ,aEor ;ene#it o# V!s and is re#erred to ;. 5ilestone S.ste,s as ;usting out o# proprietar. Eail. Choosing .our own hardware can reduce total costs and increase #le+i;ilit. to design and deplo. a s.ste, that ;est ,eets .our needs. However? ,an. V!s suppliers do o##er appliances. -ppliances in IT re#ers to ;undles o# hardware and so#tware that .ou ,ust purchase together. - cellular phone is a co,,on e+a,ple o# an appliance. Fou cannot ,i+ and ,atch phone so#tware #ro, one supplier and load it on the hardware o# another. &n the s,all scale? co,panies such as Video"rotein? o##er appliances that o##er the potential o# reducing setup and installation co,ple+it.. &n the large scale? co,panies such as Steel;o+ o##ers appliances that o##er the potential o# reducing costs and hardware necessar. #or deplo.ing 144 or 1444s o# ca,eras.
DVRs Cannot Support IP Cameras ). generall. accepted de#inition? a product re#erred to as a DV! does not support I" ca,eras. The digital in Cdigital video recorderD generall. re#ers to analog ca,era #eeds ;eing converted to digital inside o# the recorder and there#ore not ;eing sent over the I" networ8. ). de#inition? a DV! can onl. support analog inputs. There#ore? a DV! can onl. support an I" ca,era i# the video #eed #ro, the I" ca,era is #irst converted ;ac8 to analog using a <decoder.<
NVRs Support Analo" Cameras #y Enco$ers
I"Video5ar8et.In#o
1$
'ncoders are appliances that converts the video #eed #ro, an analog ca,era into an I" strea, that can ;e trans,itted over a co,puter networ8 li8e an e,ail or a CFou Tu;eD video. -l,ost all V!s support encoders. Co,,onl. held ;ene#its o# encoders include:
-llowing e+isting analog ca,eras to ;e used with V!s 'li,inating the use o# proprietar. coa+ial? twisted pair or #i;er networ8s
Some Systems are Bot% DVRs an$ NVRs (DEBA ED! So,e appliances support ;oth I" ca,eras and directl. connected analog ca,eras. Speci#icall.? these appliances do not reAuire encoders to support analog ca,eras. -nalog ca,eras can ;e directl. connected to the ;ac8 o# the appliance. This eli,inates the need #or encoders. Such appliances are generall. re#erred to a h.;rid DV!/ V!s. The ,ain ;ene#its cited #or h.;rid s.ste,s is that the. can ;e cheaper than so#tware onl. V!s and that the. ease the transition #ro, analog ca,eras to I" ca,eras. 5an. de;ate the validit. o# h.;rid s.ste,s as true V!s or I" Video Surveillance s.ste,s. 5aEor concerns include the loc8 into proprietar. hardware and the o#ten inco,plete choices o# I" ca,era support and nu,;er o# I" ca,eras a h.;rid s.ste, can support.
All NVRs Support Certain Basic &unctionalities It is widel. agreed that all V!s support certain ;asic #unctionalities:

!ecord Video View 0ive Video 24
I"Video5ar8et.In#o
Search #or !ecorded Video View !ecorded Video
Conduct these #unctionalities #ro, a re,ote co,puter
NVRs can Differ Si"nificantly in A$'ance$ &unctionalities 3hile all V!s are so#tware applications? the so#tware #unctionalities that V!s o##er can var. signi#icantl.. This variance can appear ;etween suppliers and even a,ongst supplier<s o##erings. @or instance? 5ilestone S.ste,s o##ers ( categories o# I" Video Surveillance / V! solutions and a nu,;er o# options. '+a,ples o# categories include:

)asic: s,all ca,era s.ste,s? ;asic #unctionalit. 5ediu,: ,ediu, ca,era s.ste,s? ,ore advanced ca,era and s.ste, controls 5ulti6Site: large ca,era s.ste,s with servers in ,ultiple locations Glo;al: super6large ca,era s.ste,s with #ailover and central ,anage,ent
3hile all <versions< o##er ;asics li8e video recording? viewing and searching? di##erent versions o##er ,ore power#ul tools to i,prove relia;ilit. and usa;ilit. as well as the nu,;er o# ca,eras and locations supported. 0i8ewise? signi#icant di##erences can e+ist ;etween V! suppliers in the #unctionalities? relia;ilit. and scala;ilit. the. o##er. V!s can also di##er in the t.pes o# options the. o##er. '+a,ples include:
&ptions #or Di##erent Verticals/-pplications =!etail? )an8ing? "eri,eter 21
I"Video5ar8et.In#o
"rotection>
&ptions #or Di##erent Video -nal.tics =Virtual Tripwire? 0"!? @acial !ecognition> &ptions #or -ccess Control integration? Central -lar, 5anage,ent integration? etc.
ot all suppliers will support all categories and options. So? even within V! solutions? ;u.ers ,ust e+a,ine what co,;ination o# #eatures are ,ost relevant #or the operational and securit. needs the. possess.
(ar"e an$ )rowin" Num#er of NVR Suppliers 3orldwide? there are easil. a #ew do2en suppliers o# V! solutions. That nu,;er is e+pected to grow as =1> DV! suppliers launch V! o##erings and =2> new entrants? attracted ;. growth? add o##erings.
I"Video5ar8et.In#o
22
Chapter !:
Introduction to Video C"D#Cs
Video surveillance s.ste,s do not use unco,pressed? <raw< video ;ecause o# the huge storage that unco,pressed video de,ands. Video is alwa.s co,pressed so that storage and ;andwidth costs can ;e ,ini,i2ed. C&D'Cs are a critical ele,ent o# choosing? designing and using video surveillance s.ste,s. C&D'Cs can lower the price o# overall s.ste,s and increase the usa;ilit. o# s.ste,s. -s such? having a ;asic understanding o# what a C&D'C is and wh. C&D'Cs are used is i,portant.
&un$amental Principle of CODECs The ,ost i,portant #actor to understand in video C&D'Cs is that C&D'Cs help ;alance o## di##erent costs. @or instance? let<s sa. .ou want to go to the ,all and to the super,ar8et. - #ew .ears ago? when gas was cheaper? .ou ,ight have done this in 2 separate trips. ow that gas prices have increased dra,aticall.? .ou ,ight want to co,;ine those trips. 3hat<s happening here is that as gas has ;eco,e ,ore e+pensive? .ou are willing to trade o## lower convenience #or savings in cash. 0i8ewise? using C&D'Cs is a ;alance ;etween the cost o# storage? ;andwidth and C":s. Speci#icall.: CODECs reduce the amount of bandwidth and storage needed at the expense of using more CPU cycles. -s such? selecting a C&D'C alwa.s reAuires .ou to understand the tradeo##s in cost ;etween using less ;andwidth and storage or using less C": c.cles. I"Video5ar8et.In#o 2%
Introduction to Video C&D'Cs
Generall. C": c.cles are cheaper than ;andwidth and storage so ,ore advance C&D'Cs save .ou ,one.. So,eti,es? C&D'Cs can ;e too de,anding? especiall. with ,egapi+el ca,eras and can potentiall. cost .ou ,ore in C": than .ou save in ;andwidth and storage. "lease read ,. ;asic ;andwidth tutorial #or a review o# ;andwidth<s i,pact on video surveillance.
CODECs O'er'iew Video ,ust ;e digiti2ed #or it to ;e used and viewed on a co,puter. C&D'Cs are ,eans or choices in how we ,a8e the video digital. C&D'Cs or co,pression / deco,pression technologies are used to ,odi#. the video that is ;eing digiti2ed. Si,ilar to how .ou ,ight HI" #iles on .our "C? the video is co,pressed on its wa. into the co,puter. -nd Eust li8e with opening a HI" #ile? the video is deco,pressed ;e#ore .ou use or view the video. :nli8e HI" #iles? the co,pression o# video loses so,e o# the in#or,ation =engineers re#er to this as loss. co,pression>. However? with the appropriate settings? a user cannot tell the di##erence visuall.. Iust li8e in the ,ovies or TV? video is a series o# i,ages that are displa.ed rapidl. one a#ter the other. In the :S? TV consists o# displa.ing a series o# %4 i,ages per second. 3hen we view these %4 i,ages per second? it<s CvideoD and it loo8s s,ooth. The #act that video is ,ade up o# a strea, o# i,ages is Auite i,portant #or understanding C&D'Cs. 3hen .ou use a C&D'C? .ou can co,press the video in two #unda,ental wa.s:
Co,press the individual i,age ;. itsel#
I"Video5ar8et.In#o
2(
Co,press a series o# i,ages together
3hen .ou co,press an individual i,age ;. itsel#? .ou si,pl. ta8e the i,age? run the co,pression and output the saved #ile =technicall. called intra#ra,e co,pression>. Iust li8e when .ou use 5icroso#t "aint and save as a I"'G? video co,pression o# individual i,ages wor8s Auite si,ilarl.. The di##erence with video is that .ou need to do these #or a continuous strea, o# i,ages. -s such? rather than si,pl. ;eing a I"'G? it is called 5otion I"'G or 5I"'G. The ;ene#it o# 5I"'G is that it reAuires ver. low C": use. The downside is that storage and ;andwidth use can ;e Auite high. 3hen .ou onl. co,press an individual i,age? .ou ignore what<s going on ;etween ,ultiple i,ages in a seAuence and o#ten send redundant in#or,ation. I# .ou are strea,ing video at ,ultiple #ra,es per second? .ou o#ten are sending ;asicall. the sa,e i,age over and over again. This can ;e Auite waste#ul. It<s si,ilar to so,eone calling .ou up ever. ,inute to tell .ou nothing changed. It would ;e #ar ;etter #or the person to onl. call .ou when news occurred. Fou can si,pl. assu,e during the rest o# the ti,e that the status is the sa,e. 3hen people tal8 a;out the ;ene#its o# 5"'G6( and H.2/(? not sending repetitive in#or,ation is the core source o# their strength. 've. so o#ten these C&D'Cs will send a whole i,age. The rest o# the ti,es the. onl. send updates descri;ing what parts o# the i,age have changed =technicall. called inter#ra,e co,pression>. Since it is co,,on that large parts o# the i,age re,ains the sa,e? this can result in ver. signi#icant reductions in storage and ;andwidth. @or e+a,ple? where 5I"'G ,a. send i,age a#ter i,age at 144 7)? codecs li8e 5"'G6( or H.2/( ,a. send the #irst i,age at 144 7) ;ut the ne+t % or ( i,ages at onl. 14 7) each. This approach can reduce ;andwidth and storage use ;. *4 J $4K. The downside with this approach is that it ta8es ,ore wor8 #or the co,puter to do I"Video5ar8et.In#o 2*
this. 3hen .ou are si,pl. co,pressing individual i,ages? .ou do not need to worr. a;out what happened ;e#ore or what the ne+t i,age will contain. Fou si,pl. appl. the co,pression rule and e+ecute. 3ith 5"'G6( or H.2/( .ou need to e+a,ine groups o# i,ages and ,a8e co,ple+ calculations o# what changed and what did not. Fou can i,agine this can ;eco,e ver. co,plicated and consu,e lots o# C": resources. H.2/( and 5"'G6( are si,ilar in that the. ;oth reduce ;andwidth and storage ;. e+a,ining groups o# i,ages when the. co,press video. - 8e. di##erence with H.2/( is that it uses ,uch ,ore co,ple+ and sophisticated rules to do the co,pression. )ecause H.2/(<s rules are ,ore sophisticated? the. can reduce ;andwidth and storage even ,ore than 5"'G6(. However? the trade6o## is that it ta8es ,ore C": c.cles to do it.
(oo*in" at Current Vi$eo Sur'eillance Systems The general trend in video surveillance has ;een a continuous ,ove,ent to C&D'Cs that save ;andwidth and storage. Historicall.? .ou have seen products ,ove #ro, 5I"'G to 5"'G6( to H.2/(. The reason wh. this has happened is ;ecause the cost o# C":s to co,press the video has decreased #aster than the cost o# ;andwidth and storage. 5ost e+perts e+pect this trend to continue. !ecentl.? the ;iggest challenge using C&D'Cs in video surveillance s.ste,s has occurred with the rise in ,egapi+el ca,eras. @or .ears? the ,a+i,u, resolution o# securit. ca,eras was constant. -ll o# a sudden with ,egapi+el ca,eras? the resolution o# securit. ca,eras has increased ;. (44K to *444K or ,ore. The greater the resolution? the harder the C": needs to wor8 and the ,ore c.cles that need to ;e allocated. The huge increase in resolution is a little si,ilar to the Eu,p in gas prices. It has
I"Video5ar8et.In#o
2/
changed the econo,ics o# C&D'Cs. 3hereas historicall.? #or standard de#inition securit. ca,eras? C": c.cles were cheaper than ;andwidth and storage. ow? since so ,uch ,ore C": c.cles are needed? it can cost wa. ,ore in C": than what .ou save in ;andwidth and storage. -s such? al,ost all co,,ercial ,egapi+el ca,eras use 5I"'G. &ne o# the ,ost i,portant ele,ents in the ne+t #ew .ears will ;e the develop,ent o# new approaches and use o# new C":s to reduce the cost o# using H.2/( #or ,egapi+el ca,eras. 5uch li8e alternative energ. develop,ent hopes to ;ring the cost o# energ. down? new approaches are ;eing sought to reduce the use o# C": c.cles in co,pressing ,egapi+el ca,era #eeds.
Conclusion :nderstanding the ;asic choices in C&D'Cs and rationale #or choosing C&D'Cs is a 8e. ele,ent in video surveillance s.ste,s. Securit. ,angers should consider:
The video Aualit. o# a ,anu#acturer<s video =this can var. widel. depending on the codecs? the settings and the i,ple,entation>. Generall. #avor the use o# a ,ore power#ul C&D'C as it can save hundreds o# dollars per ca,era in reduced storage costs.
I"Video5ar8et.In#o
21
Chapter $:
%andwidth %asics or Video Surveillance
3hen using I" ca,eras? 5egapi+el ca,eras? V!s or even DV!s? understanding the ;asics a;out how ,uch ;andwidth is availa;le and how ,uch is needed is critical in planning? designing and deplo.ing I" video surveillance s.ste,s. 'ver.one in the industr. should have an understanding o# the ;asics as ;andwidth is a critical #actor in video surveillance
+ow Muc% Ban$wi$t% is A'aila#le, To #igure out how ,uch ;andwidth is availa;le? .ou #irst need to deter,ine what locations .ou are co,,unicating ;etween. 5uch li8e driving? .ou will have a starting point and destination. @or e+a,ple? #ro, .our ;ranch o##ice to .our headAuarters. However? unli8e driving? the a,ount o# ;andwidth availa;le can range dra,aticall. depending on where .ou are going. The ,ost i,portant #actor in deter,ining how ,uch ;andwidth is availa;le is whether or not .ou need connectivit. ;etween two di##erent ;uildings. @or instance: Ban$wi$t% )enerally A'aila#le Sa,e )uilding Di##erent )uildings 145;/s to 144 5;/s .* 5;/s to * 5;/s o#
The a,ount o# ;andwidth availa;le going #ro, .our o##ice to a co6wor8er<s o##ice in the sa,e ;uilding can ;e 244 ti,es ,ore than the ;andwidth #ro, .our o##ice I"Video5ar8et.In#o 28
)andwidth )asics #or Video Surveillance
to a ;ranch o##ice down the ;loc8. This is true in $4K or ,ore cases. 5ore ;andwidth ,a. ;e availa;le in the #ollowing conditions:

Di##erent ;uildings ;ut on the sa,e ca,pus In a central ;usiness district o# a ,aEor cit. Fou are a teleco,,unications or research co,pan.
Different Buil$in"s The 8e. driver in ;andwidth availa;ilit. is the cost o# deplo.ing networ8s ;etween ;uildings. Generall. re#erred to as the 3ide -rea etwor8 or 3- ? this t.pe o# ;andwidth is usuall. provided ;. teleco,,unications co,panies. &ne co,,on e+a,ple is ca;le ,ode, or DS0? which can provide an.where #ro, .* 5;/s to * 5;/s at L*4 to L1*4 per ,onth. -nother e+a,ple is a T1? which provides 1.*5;/s #or a;out L%44 to L/44 per ,onth. -;ove this level? ;andwidth generall. ;eco,es ver. e+pensive. In ,ost locations? getting 145;/s o# ;andwidth can cost thousands per ,onth. 5an. tal8 a;out #i;er =so,eti,es called @TTH/@TTC> ;ut #i;er to the ;uilding is not and will not ;e widel. availa;le #or .ears. @i;er to the ho,e or to the ;usiness pro,ises to reduce the cost o# ;andwidth signi#icantl.. evertheless? it is ver. e+pensive to deplo. and despite e+cited discussions #or the last decade or ,ore? progress re,ains slow. I# .ou have it great? ;ut do not assu,e it.
Same Buil$in"s ). contrast? ;andwidth inside o# ;uildings =or ca,puses> is Auite high ;ecause the costs o# deplo.ing it are Auite low. on technical users can easil. set up a 14445;/s networ8s inside a ;uilding =a8a 0ocal -rea etwor8s or 0- s> #or less I"Video5ar8et.In#o 2$
than L1?444 installation cost with no ,onthl. costs. Contrast this to the 3- ? where the sa,e ;andwidth could cost tens o# thousands o# dollars per ,onth. The cost o# deplo.ing networ8s in ;uildings are low ;ecause there are ,ini,al to no construction e+penses. 3hen .ou are ;uilding a networ8 across a cit.? .ou need to get rights o# wa.s? trench? install on telephone poles? etc. These are ,assive proEects that can easil. de,and ,illions or ;illions o# dollars in up #ront e+penses. ). contrast? inside a ;uilding? the ca;les can o#ten ;e Auic8l. and si,pl. #ished through ceilings =not the pro#essional wa. to do it ;ut the wa. ,an. people do it in deplo.,ents>. - lot o# discussion a;out wireless =3i5a+? 3i@i? %G? etc> e+ists ;ut wireless will not provide signi#icantl. greater ;andwidth nor signi#icantl. ;etter costs than DS0 or ca;le ,ode,. -s such? wireless will not solve the e+pense and li,itations o# ;andwidth ;etween ;uildings. That ;eing said? wireless a;solutel. has ;ene#its #or ,o;ilit. purposes and connecting to re,ote locations that DS0 or ca;le ,ode, cannot cost e##ectivel. serve. The point here is si,pl. that it will not solve the pro;le, o# ;andwidth ;etween ;uildings ;eing ,uch ,ore e+pensive than ;andwidth inside o# ;uildings.
+ow Muc% Ban$wi$t% Do IP Cameras Consume, @or the ;andwidth consu,ption o# an I" ca,era? use 1 5;/s as a rough rule o# thu,;. ow? there are ,an. #actors that a##ect total ;andwidth consu,ption. Fou can certainl. strea, an I" ca,era as low as .2 5;/s =or 244 7;/s> and others as high as / 5;/s. The ,ore resolution and greater #ra,e rate .ou want? the ,ore ;andwidth will ;e used. The ,ore e##icient the C&D'C .ou use? the less ;andwidth will ;e used. @or the ;andwidth consu,ption o# a 5egapi+el ca,era? use * 5;/s to 14 5;/s as
I"Video5ar8et.In#o
%4
a rough rule o# thu,;. -gain? there are a nu,;er o# #actors that a##ect total ;andwidth consu,ption. - 1.% ,egapi+el ca,era at 1#ps can consu,e as little as .8 5;/s =or 844 7;/s> .et a * ,egapi+el ca,era can consu,e as ,uch as (* 5;/s.
-%at Does t%is Mean for my IP Vi$eo System, Iust li8e dealing with personal #inance? we can now #igure out what we can <a##ord<: .Ban$wi$t% Bu$"et A'aila#le/ )etween )uildings Inside )uildings .* 5;/s to * 5;/s 14 5;/s to 144 5;/s
.Ban$wi$t% Cost/ I" ca,eras 5egapi+el ca,eras 1 5;/s * 5;/s to 14 5;/s
:sing this chart? we can Auic8l. see what co,;ination o# I" and ,egapi+el ca,eras we can use ;etween ;uildings or inside o# ;uildings. 1. Inside o# ;uildings? it is eas. to strea, nu,erous I" and ,egapi+el ca,eras. 2. )etween ;uildings? it is al,ost i,possi;le to strea, nu,erous I" and ,egapi+el ca,eras. )ecause o# this situation? the standard con#iguration one sees in I" Video s.ste,s is: I"Video5ar8et.In#o %1
- local recorder at each ;uilding/re,ote site. The local recorder receives the strea,s #ro, the ;uilding and stores the,. The local recorder onl. #orwards the strea,s =live or recorded> o##6site when a user speci#icall. wants to view video. !ather than overloading the 3- networ8 with unrealistic ;andwidth de,ands all da. long? ;andwidth is onl. consu,ed when a user wants to watch. Generall.? re,ote viewing is sporadic and I" video coe+ists nicel. with the e+pensive 3ide -rea etwor8. The local recorder has ;uilt6in #eatures to reduce the ;andwidth needed to strea, video to re,ote clients. 5ost s.ste,s have the a;ilit. to reduce the #ra,e rate o# the live video strea, or to d.na,icall. reduce the video Aualit. to ensure that the video s.ste, does not overload the networ8 and that re,ote viewers can actuall. see what is going on the other side. Generall.? the live video strea, is su##icient to identi#. the ;asic threat. In an. event? ;andwidth is generall. so costl.? especiall. the upstrea, ;andwidth needed to send to a re,ote viewer? that this is the ;est #inancial decision.
Conclusion 7nowing how ,uch ;andwidth is availa;le #or DV!s and V!s and how ,uch ;andwidth I" and ,egapi+el ca,eras consu,e are 8e. ele,ents in planning and deplo.ing via;le I" video s.ste,s. Though this is si,pl. a ;road surve.? ,. hope is that this helps identi#. #unda,ental ele,ents in understanding the i,pact o# ;andwidth on I" video.
I"Video5ar8et.In#o
%2
Chapter &:
#'a(ining Video )nal*tics
@or * .ears now? the pro,ise o# using video anal.tics to stop trespassers crossing #ences? catch thieves in stores? detect a;andoned o;Eects? etc has ;een a #reAuent topic o# discussion. 3hile video anal.tics holds great pro,ise? people are still as8ing a;out the via;ilit. o# using anal.tics in the real world. Indeed? as stories o# video anal.tic pro;le,s have spread? concerns a;out the ris8s o# video anal.tics now see, higher than a #ew .ears ago when the novelt. o# the technolog. spurred wide e+cite,ent. This article surve.s the ,ain pro;le,s li,iting the use and growth o# video anal.tics. It is ,eant to help securit. ,anagers gain a ;etter sense o# the core issues involved. Top % "ro;le,s: 1. 'li,inating @alse -lerts 2. S.ste, 5aintenance Too Di##icult %. Cost o# S.ste, Too High Eliminatin" &alse Alerts Since the goal o# video anal.tics is to eli,inate hu,an involve,ent? eli,inating #alse alerts is necessar. to acco,plish this. 'ach #alse alerts not onl. reAuires a hu,an assess,ent? it increases e,otional and organi2ational #rustration with the s.ste,. I"Video5ar8et.In#o %%
'+a,ining Video -nal.tics
5ost are #a,iliar with ;urglar alar, #alse alar,s and the #rustration these causes. &n average? ;urglar alar, #alse alar, per house or ;usiness are #airl. rare. I# .ou have 1 or 2 per ,onth? that is #airl. high. 5an. people do not e+perience #alse alar,s o# their ;urglar s.ste, #or ,onths. ). contrast? ,an. video anal.tic s.ste,s can generate do2ens o# #alse alar,s per da.. This creates a #ar greater issue than an.thing one is accusto,ed to with ;urglar alar,s. "lus? with such alar,s happening ,an. ti,es throughout the da.? it can ;eco,e an operational ;urden. ow? not all video anal.tics s.ste,s generate lots o# #alse alar,s ;ut ,an. do. These issues have ;een the nu,;er one issue li,itation o# the integrators and end6 users that I 8now using and tr.ing video anal.tics.
System Maintenance oo Difficult S.ste, ,aintenance is a o#ten overloo8ed and so,ewhat hidden issue in video anal.tics. &ver a period o# wee8s or ,onths? a video anal.tic s.ste,<s #alse alerts can start rising considera;l. due to changes in the environ,ent? weather and the position o# the sun. This can suddenl. and surprisingl. cause ,aEor pro;le,s with the s.ste,. ot onl. is the increase in #alse alerts a pro;le,? the ris8 now that the s.ste, could une+pectedl. ;rea8 in the #uture creates a signi#icant pro;le, in trust. I# .our peri,eter surveillance one da. stops #unctioning properl.? .ou now have a serious #law in .our overall securit. plan.
I"Video5ar8et.In#o
%(
This has ;een a cause o# a nu,;er o# video anal.tic s.ste, #ailures. The s.ste,s? alread. purchased? si,pl. get put to the site ;eco,ing a ver. e+pensive testa,ent to not ;u.ing or re#erring one<s colleagues to video anal.tics. This ;eing said? not all video anal.tic s.ste,s e+hi;it this ;ehavior ;ut .ou would ;e prudent to care#ull. chec8 re#erences to veri#. that e+isting s.ste,s have ;een operating #or a long period o# ti,e without an. ,aEor degradation.
Cost of System oo +i"% 3hile .ou can #ind ine+pensive video anal.tic s.ste,s toda.? these s.ste, tend to e+hi;it pro;le,s 1 and 2? high #alse alerts and poor s.ste, ,aintenance. Indeed? in ,. e+perience? video anal.tic s.ste,s that are either #ree or onl. cost L1446L244 ,ore generall. have signi#icant operational pro;le,s. &ne co,,on #eature o# s.ste,s that wor8 is that the co,plete price #or hardware and so#tware is usuall. L*44 or ,ore per channel #or the anal.tics. ow Eust ;ecause a video anal.tic s.ste,s is e+pensive o;viousl. does not ,ean it is good. However? there are necessar. costs in ;uilding a s.ste,s that is ro;ust and wor8s well in the real world. The cost o# video anal.tic s.ste,s co,es in ,a8ing the, ro;ust to real world conditions that we all ta8e #or granted. The developer needs to ,a8e the video anal.tic s.ste, CintelligentD enough to handle di##erences in lighting? depth? position o# the sun? weather? etc. Doing this involves ;uilding ,ore co,ple+ or sophisticated progra,s. Such progra,s al,ost alwa.s reAuire signi#icantl. ,ore co,puting hardware to e+ecute and signi#icant ,ore capital invest,ent in writing? testing and opti,i2ing the progra,. -ll o# these clearl. increase costs. The challenge is that it is ;asicall. i,possi;le to see this #ro, ,ar8eting
I"Video5ar8et.In#o
%*
de,onstrations ;ecause #ro, a de,o all s.ste,s invaria;l. loo8 e+actl. ali8e. This o# course has the vicious e##ect o# encouraging people to choose cheaper s.ste,s that are ,ore li8el. to generate high #alse alerts and ;e un,aintaina;le. I# .ou select a s.ste, that wor8s? the cost per ca,era can ,a8e it di##icult to Eusti#. the e+pense. Indeed? so ,uch o# the #irst generation video anal.tic deplo.,ents? ca,e #ro, govern,ent grant ,one.? essentiall. ,a8ing the cost secondar. or not relevant. evertheless? #or video anal.tics to grow in the private sector? the. will not onl. need to wor8 the. will need to generate #inancial return. 3hen video anal.tics allow #or guard reduction or reduce high value #reAuent losses? it is eas. to Eusti#. and .ou see co,panies having success here =in ter,s o# pu;licl. docu,ented cases? IoI,age is the leader here>. @or other cases? where hu,ans are not ;eing eli,inated? the individual loss is s,all or the occurrence o# loss is low? the cost can ;e a ,aEor ;arrier.
Conclusion Though I anticipate video anal.tics successes to increase? I ;elieve such success will ;e constrained to applications where the loss characteristics and/or the hu,an reduction costs are high. 3hile anal.tics will certainl. ;eco,e cheaper? such cost decreases will ta8e ti,e and in the interi,? it is these high value applications where anal.tics can gain a #oothold o# success. )oth testing and re#erence testing are critical to the use o# video anal.tics.
I"Video5ar8et.In#o
%/
Chapter +:
,icense Plate Recognition -utorial
0icense "late !ecognition is perhaps the ,ost ,ature and read. to use video anal.tic availa;le #or securit. ,anagers toda.. evertheless? 0"! is a ver. de,anding application that can onl. succeed in li,ited operational conditions deplo.ed ;. e+pert securit. integrators. Historicall.? pu;licl. availa;le in#or,ation clearl. e+plaining the operational i,pact has ;een hard to #ind. Than8#ull.? 5ilestone has released their 0"! ad,inistrator<s ,anual providing an honest? clear and concise e+planation. I reco,,end .ou read pages 2$6%* to get a ver. rapid ;ut deep review o# the 8e. #actors. Though this is #or 5ilestone the points are generall. consistent with the state o# the art in currentl. availa;le co,,ercial s.ste,s. The 5ilestone docu,ent helps to reveal % 8e. practical ele,ents: 1. 0"! can onl. succeed when a nu,;er o# strict operational conditions are ,et. 2. The costs o# achieving these conditions ,a8es 0"! un#easi;le #or ,an. scenarios. %. Fou need deep securit. integration e+pertise to succeed ;ut onl. ,odest IT depth.
%e Con$itions Here are the 8e. conditions that need to ;e ,eet in appro+i,ate order o# di##icult.:
I"Video5ar8et.In#o
%1
0icense "late !ecognition Tutorial
1. :S license plates need to ;e at least 1%4 pi+els wide. This translates roughl. into an i,age no wider than *6/ #eet assu,ing (CI@ standard de#inition video. That<s a ver. tight shot. 2. The hori2ontal angle ;etween the ca,era and plate is within 24 degrees. This ,eans that i# .our ca,era is 14 #eet awa. #ro, the plate? the plate cannot ;e ,ore than % #eet to the right or le#t o# the ca,era. This signi#icantl. li,its where .ou can put the ca,era. %. The vertical angle ;etween the ca,era and plate is within %4 degrees. This ,eans that i# .our ca,era is 14 #eet awa. #ro, the plate and the plate is % #eet o## the ground? the ca,era cannot ;e ,ounted ,ore than 8 #eet high. This usuall. can ;e acco,,odated ;ut is low relative to nor,al heights #or outdoor surveillance. (. There are a host o# lighting adEust,ents that need to ;e ,ade. Si,pl. using a stoc8 ca,era with stoc8 settings will routinel. cause ver. poor per#or,ance. @or e+a,ple? 5ilestone reco,,ends C5&S ca,eras? disa;ling auto gain? using 3D! and higher shutter speeds =i# the car is ,oving>. There is a lot o# advanced details that need to ;e set correctl.. -l,ost all success#ul 0"! deplo.,ents used speciali2ed ca,eras? o#ten ,anu#actured speci#icall. #or license plate rate recognitions. These ca,eras generall. have ;uilt in in#rared illu,inators and pre6set con#iguration opti,i2ed #or #iltering out lighting issues. *. Fou ,ust use 5I"'G and .ou cannot use H.2/( or 5"'G6(. Since the anal.tics in this design are ;eing done outside o# the ca,era and since the anal.tic can onl. process i,ages? 5I"'G is reAuired. Fou could theoreticall. use H.2/( or 5"'G6( ;ut then .ou would have to decode it and the processing power can ;e ver. signi#icant. )otto, line is this can have a ;ig i,pact on ;andwidth utili2ation especiall. i# .ou are loo8ing #or a wireless s.ste,.
I"Video5ar8et.In#o
%8
&easi#ility Clearl.? 0"! is #easi;le #or the traditional license plate ca,era use case: - ca,era installed i,,ediatel. adEacent to an entrance or toll ;ooth that is onl. a #ew #eet o## the ground and dedicated to loo8ing at the plate. -uto,ated 0"! ,a8es reading these plates easier. However? #or ;roader ,ar8et usage? this has ,aEor li,itations. 0ots o# co,panies li8e the concept o# ,onitoring the license plates o# people who enter their pre,ises. Setting up ca,eras in the speci#ic constraints reAuired can ;e ver. e+pensive. -ssu,ing .ou can #ind a location that ,eets these constraints? it reAuires a construction proEect that can ;e L*?444 or ,ore per ca,era si,pl. #or the installation and eAuip,ent. The hol. grail is reutili2ing .our "THs ,ounted on roo#s and poles. However? these conditions should ,a8e it clear that is not #easi;le. &ne? getting the resolution needed would ;e di##icult. Does a ,onitor ,anuall. 2oo, in on license plates9 'ven i# he does? what will the i,age Aualit. ;e? given the lighting constraints reAuired #or 0"!. -lso? it will ;e e+tre,el. tough to sta. within ;oth the hori2ontal and vertical angle reAuire,ents. 0"! anal.sis? with its current capa;ilities? cannot ena;le signi#icantl. new operational uses o# license plate ,onitoring. 3hile it should help with the traditional use case o# ,onitoring controlled tra##ic #low? its constraints ,a8e it ver. challenging #or ;roader use. Security Inte"ration E0pertise The other interesting ele,ent that the 5ilestone ,anual de,onstrates is that 0"! integration does not de,and deep IT s8ill ;ut it does de,and deep e+pertise in
I"Video5ar8et.In#o
%$
securit. design and ca,era s.ste,s. Integrating 0"! is ,uch ,ore li8e using a graphics design application than it is li8e setting up a ,ail server. It depends on understanding the design o;Eectives o# securit.? the ph.sical conditions o# the site and the capa;ilities o# the video tools availa;le. The IT ele,ents o# the setup are #airl. straight#orward #or a securit. integrator. The challenge lies in the design and application.
I"Video5ar8et.In#o
(4
Chapter .:
Introduction to DVR/NVR Storage "pti(i/ation
Storage opti,i2ation is a ,aEor concern #or securit. ,anagers as storage costs #or video surveillance have alwa.s ;een a large portion o# the overall purchase price. 5egapi+el ca,eras have ;rought renewed interest in ,easures to ,a+i,i2e V!/ DV! storage duration and use. Cost is a ;ig #actor as the potential storage needed could increase ;. 14+ or ,ore than historical standards. :nderstanding what options and ,easures are availa;le is ;eco,ing increasingl. i,portant to selecting V!s/DV!s and designing I" video s.ste,s. This report surve.s co,,on ,easures used to ,a+i,i2e strorage duration and use. !ecentl.? interest has risen in new product categories that speciali2e in opti,i2ing storage use. @rost and Sullivan has recentl. reviewed BVideo 0i#ec.cle 5anage,ent SolutionsB and identi#ied Ti,eSight S.ste,s as a B.oung leaderB in this space. The release does a good Eo; o# identi#.ing the pro;le, and highlighting one potential solution. -s al,ost all video s.ste,s have nu,erous ,easures to opti,i2e storage use? I reco,,end that integrators and end users #ocus on utili2ing e+isting ,easures in leading s.ste,s. Video s.ste, developers have ;een ;uilding tools #or .ears to address storage opti,i2ation. 5ost will ;e ;est served ;. selecting a video ,anage,ent s.ste, ;ased on #eatures opti,i2ed #or .our speci#ic securit. needs. Signi#icant and co,para;le storage opti,i2ation can generall. ;e acco,plished on ,ost ,ainstrea, V! / DV! s.ste,s.
+ow Do I Optimi1e Stora"e,
I"Video5ar8et.In#o
(1
Introduction to DV!/ V! Storage &pti,i2ation
This report reviews 8 co,,onl. availa;le storage opti,i2ation #unctions availa;le on ,ainstrea, V! /DV! s.ste,s. Though not ever. s.ste, has all o# these #eatures? all s.ste,s o##er a nu,;er o# the,? providing strong storage opti,i2ation. Here is the list:

)asic 5otion -nal.tics -dvanced Video -nal.tics 5otion '+clusion Hones Data -ging !ecording Schedule C&D'C Selection Dual Strea,ing Storage Clusters
A$'ance$ Video -nal.tics ow that video anal.tics are getting accurate at detecting people? #aces and vehicles? this intelligence can ;e used to control recording. I ;elieve this will ;eco,e one o# the ,ost power#ul new areas o# storage opti,i2ation in the ne+t % .ears. 0ong ter, storage can ;e opti,i2ed ;. selectivel. recording o;Eects ,ost li8el. to ;e o# long ter, interest 6 people? #aces and vehicles. Traditionall.? long ter, storage opti,i2ation techniAues reduce the Aualit. or the #ra,e rate o# all video uni#or,l.. 3ith video anal.tics? storage opti,i2ation techniAues can ;eco,e s,arter? increasing the pro;a;ilit. o# possessing Aualit. long ter, evidence while ,ini,i2ing total storage consu,ed. @or instance? in addition to recording video? %V! records all #aces seen on ca,eras. @aces o# all the people =144?444M> conducting transactions at a ;an8 I"Video5ar8et.In#o (2
;ranch can ;e stored at (CI@ Aualit. with less than 24G) o# storage. This is 1/144th the a,ount o# storage needed #or video and the ,ost i,portant evidence #or retail ;an8<s securit. needs. &# course? toda. this is Eust #aces ;ut the sa,e process can and will certainl. eventuall. ;e used to store all the people seen? all the cars ,oving through an area? etc. Video anal.tic co,panies speciali2ing in peri,eter violation are reducing storage needs #or those ca,eras ;. $4K or ,ore. ). placing intelligence in the ca,era? the ca,era can onl. strea, or the ,anage,ent s.ste, can onl. record speci#ic o;Eects o# interest. @or ca,eras whose ,ain purpose is real ti,e alerting? this is a great storage win. &# course? ,an. ca,eras are needed #or investigation purposes and need storage. -s such? this is si,pl. another tool in our collection.
Basic Motion Analytics 5ost video surveillance deplo.,ents use ;asic ,otion anal.tics to control recording. )ecause ,ost #acilities have signi#icant periods o# low activit. =e.g.? nights? wee8ends> and areas o# low activit. =e.g.? hallwa.s? stairwells>? ,otion anal.tics can reduce storage consu,ption ;. *4K to 84K. 5ost s.ste,s set their ;asic ,otion anal.tics to ;e #airl. conservative so that the. rarel. ,iss real incidents. -s such? ;asic ,otion anal.tics is trusted and used ;. ,an. ,ilitar. ;ases? ;an8s and @ortune 144 co,panies and ,ost real world deplo.,ents. &# course? so,e #acilities do not want to ta8e an. ris8 and reAuire continuous recording. - nice ;alance that is so,eti,es achieved is a co,;ination o# continuous and ,otion ;ased recording with a ;aseline level o# continuous recording =e.g.? % #ra,es per second> and ,otion ;ased recording set higher =sa. to 1* #ps>. This ensures that video is alwa.s recorded ;ut storage use is opti,i2ed #or when activit. o# interest is ,ost li8el. to occur =that is? when ,otion is detected>.
I"Video5ar8et.In#o
(%
Motion E0clusion 2ones :sing ;asic ,otion anal.tics to control recording is enhanced through using ,otion e+clusion 2ones. It is co,,on #or ca,eras to cover areas that are not o# interest to users. '+a,ples include highwa.s ;ehind the ;uilding? a tree out #ront? windows? ceiling lights? etc. Ta8ing a #ew ,inutes to set up ,otion e+clusion 2ones can reduce the storage utili2ation ;. up to *4K on certain ca,eras. -#ter the #irst wee8 o# a new install? a review should ;e conducted to tune these settings.
Data A"in" 5an. s.ste,s reduce the nu,;er o# #ra,es in stored video as the video is older. The ;asic pre,ise is the older the video? the lower the pro;a;ilit. that the video is relevant. !ather than si,pl. delete the video? the si2e o# the video is reduced so that so,e evidence is availa;le Eust in case ;ut the storage costs are ,ini,i2ed. @or instance? 5arch etwor8s has a #eature called BIntelligent Video !etention.B -vigilon has an advanced data aging solution that speciali2es in opti,i2ing storage #or ,ulti6,egapi+el ca,eras. In higher end video s.ste,s? this t.pe o# #eature is #reAuentl. availa;le. It<s Auite use#ul ;ecause it can easil. dou;le storage duration.
Recor$in" on a Sc%e$ule 5an. organi2ations have greater securit. ris8s at di##erent ti,es o# the da.. Schedules are a co,,on #eature to adEust recording para,eters to ,atch those
I"Video5ar8et.In#o
((
di##erent level o# ris8s. @or instance? an organi2ation ,a. want continuous recording during ;usiness hours ;ut is o8 with onl. having ,otion ;ased recording a#ter hours. 5a8ing this adEust,ent can reduce video storage use ;. up to (4K.
CODEC C%oice Choosing a video C&D'C that provides the ,ost e##icient storage utili2ation has ;een a 8e. co,ponent o# video s.ste, designs #or .ears. 3hile technical issues e+ists? the trend o# ,oving #ro, less e##icient to ,ore e##icient C&D'Cs is clear =e.g.? #ro, 5I"'G to 5"'G6( to H.2/(>. The 8e. practical issue currentl. is the use o# H.2/( #or ,egapi+el ca,eras due to the high s.ste, reAuire,ents H.2/( de,ands. 3ith ,ultiple ,egapi+el ,anu#acturers releasing H.2/( ,egapi+el ca,eras? in the ne+t #ew .ears H.2/( ,egapi+el ca,eras loo8s certain to ;e a realit. =at least #or lower resolution 5" ca,eras>. 5igrating #ro, 5I"'G to H.2/( can reduce storage use ;. *4K or ,ore.
Dual Streamin" To ,a+i,i2e C&D'Cs di##erent strengths and wea8nesses? ,ultiple video strea,s can ;e used. @or instance? H.2/( ,a. ;e the ;est choice #or storage opti,i2ation ;ut 5I"'G has advantages o# live ,onitoring =e.g.? lower dela.? lower processing power to view>. 5ost ca,eras support dual strea,ing. Video surveillance s.ste,s can ta8e advantage o# this to reduce storage costs while ensuring opti,al live video ,onitoring.
Stora"e Clusters
I"Video5ar8et.In#o
(*
Historicall.? storage was separated into s,all pools #or each unit and options #or upgrading storage were li,ited to a #ew T)s =at ,ost>. Toda.? with storage clusters #or video surveillance ,aturing? centrali2ed pools o# storage can create N %4K e##iciencies in storage use and ,a8e e+tending storage si,ple and #airl. ine+pensive. I# .ou are interested in learning ,ore? read ,. e+tensive anal.sis o# storage clusters.
Conclusion 3hile ;. no ,eans co,prehensive? this surve. should help engineers and users to identi#. and use co,,onl. availa;le ,easures to opti,i2e storage duration. :nderstand what .our s.ste,s o##er and ,a8e use o# those #eatures. ). doing so? .ou will ;e a;le to opti,i2e the storage o# ,ost an. DV! or V! and acco,,odate increasing storage de,ands.
I"Video5ar8et.In#o
(/
Chapter 0:
1ireless Video Surveillance -utorial
3hile wireless can uniAuel. solve certain challenges? it is #ar ris8ier to deplo. and use than wired networ8s. -s such? it is critical to understand when to use wireless s.ste,s and the 8e. ris8s in designing such s.ste,s. I# .ou use wireless networ8s prudentl. #or video surveillance s.ste,s? the #inancial ;ene#its can ;e Auite signi#icant. However? ,iscalculation in choice and design can result in signi#icant relia;ilit. and scala;ilit. pro;le,s. -s a general rule? .ou should avoid using wireless networ8s unless wired networ8s costs are signi#icantl. higher than a wireless s.ste,. This is ;ecause deplo.ing and ,aintaining wireless networ8s is #ar ,ore ris8. and e+pensive than it is #or a wired networ8. 3ireless s.ste,s #ace ,uch ,ore serious pro;le,s that wired networ8s do such as constrained ;andwidth? signal o;struction? higher ,aintenance cost and scala;ilit. restrictions. 0et<s review these 8e. ele,ents: 6 How ,uch ;andwidth is availa;le9 6 How #ar awa. can the wireless ca,eras ;e9 6 How ,an. ca,eras can I deplo.9
Ban$wi$t% 3ireless networ8s have #ar lower ;andwidth than wired networ8s. &n a wired networ8? ;andwidth availa;le #or video surveillance can ;e easil. 14 5;/s to 144 5;/s. &n a wireless networ8? .our availa;le ;andwidth is o#ten no ,ore than * I"Video5ar8et.In#o (1
3ireless Video Surveillance Tutorial
5;/s to 2* 5;/s. It is a dra,atic and o#ten overloo8ed aspect o# wireless video surveillance design. 3ireless video surveillance usuall. has signi#icantl. less ;andwidth the wireless s.ste, states. This is ;ecause the wa. ;andwidth is calculated in wireless s.ste,s is the opposite o# the ,ore traditional wired approach. 3ith a wired networ8? i# .ou sa. .ou have 144 5;/s ;andwidth? this ,eans .ou have 144 5;/s going up and another 144 5;/s going down. In a wireless networ8? i# .ou sa. .ou have 11 5;/s ;andwidth? that is the total #or ;oth upstrea, and downstrea,. So,e wireless s.ste,s are #i+ed to allow hal# the ;andwidth #or upstrea, and hal# #or downstrea,. This is a ;ig pro;le, #or video surveillance ;ecause al,ost all the ;andwidth used is in one direction =upstrea,>. 5a8e sure .our wireless s.ste, lets the upstrea, ta8e up the whole ;andwidth i# needed. This is co,,on with wireless s.ste,s dedicated to video ;ut none in co,,on co,,ercial gear. 'nviron,ental conditions o#ten reduce the ;andwidth #urther. 3ireless networ8s are ,uch ,ore prone to e##ects #ro, the environ,ent than wired networ8s. 3ireless networ8s will onl. achieve their ,a+i,u, i# the strength o# the signal =signal to noise> is su##icientl. high. I# there are partial o;structions or i# the antenna shi#ts slightl.? the ;andwidth #ro, wireless s.ste,s can drop #urther. In our previous e+a,ple? the 11 5;/s wireless s.ste, onl. o##ers *.* 5;/s #or strea,ing video. However? co,,on environ,ental conditions can drop the ;andwidth to 2.1* 5;/s.
Distance of Cameras It is Auite hard to set up ,ulti6,ile wireless lin8s to video surveillance ca,eras. nu,;er o# #actors including o;structions? #reAuenc. li,itations? power li,itations? and installation precision drive this. ote: this tutorial assu,es the use o# unlicensed #reAuenc.? ;. #ar the ,ost co,,on choice #or deplo.ing
I"Video5ar8et.In#o
(8
wireless video s.ste,s. I# .ou are using licensed #reAuenc.? where .ou can use ,uch higher power and ensure no inter#erence? these issues are not as signi#icant. However? o;taining licenses are e+pensive and ti,e consu,ing so ,ost application use unlicensed spectru,. The rest o# the discussion assu,es unlicensed #reAuencies. Fou are constrained in how power#ul .our signal can ;e? signi#icantl. reducing the distance that .ou can trans,it. The govern,ent restricts the power o# .our signal so that .ou do not drain out other users. However? this ,eans it is ,uch harder to push through o;stacles and go greater distances. It also ,eans that other users o# the sa,e #reAuenc. can reduce the ;andwidth or ;loc8 .our signal. This is a ,aEor #actor in the e,ergence o# the (.$ Gh2 range #or use in video surveillance proEects as that range is dedicated to pu;lic sa#et.. &;stacles are ver. seriousl. pro;le,s #or wireless video surveillance s.ste,s. 5ost wireless video surveillance s.ste, use #reAuenc. ranges that are easil. a;sor;ed ;. ;uildings and trees =2.( Gh2 through *.8 Gh2>. "racticall. spea8ing? .ou ,a. want to trans,it to a ;uilding 144 ,eters awa. ;ut i# another ;uilding is in ;etween? the signal will ;e a;sor;ed and the lin8 will not ;e possi;le. Fou can and should use ,esh networ8s to acco,,odate this ;ut .ou ,ust #actor in the i,pact on the cost o# the overall networ8. Installation precision is 8e. ;ut issues can go wrong that ,a. increase long ter, ,aintenance. )ecause o# power restrictions? wireless video s.ste,s co,,onl. use high gain antennas that increase signal power ;. concentrating it into a narrower area. This can help greatl. in going longer distances or overco,ing o;stacles? however? it ,eans the antennas ,ust line up ver. precisel.. I# the. do not? the per#or,ance o# the s.ste, will degrade signi#icantl.. -lso? i# during the li#e o# the s.ste,? either antenna shi#ts? the per#or,ance o# the s.ste, could degrade <out o# the ;lue.<
I"Video5ar8et.In#o
($
Num#er of Cameras The nu,;er o# ca,eras on a wireless s.ste, is severel. constrained due to ;andwidth li,itations and constraints on how #ar ca,eras can ;e placed. @or an. given wireless connection? the ,a+i,u, nu,;er o# ca,eras that can ;e supported is generall. ;etween * and 1* with the ca,eras ;eing less than a ,ile #ro, the receiver. 'ven <VC!< Aualit. video using a good C&D'C will ta8e a;out 1 5;/s. This is signi#icant when .our are dealing with wireless lin8s that ,a. onl. support * 6 24 5;/s. The total nu,;er o# wireless ca,eras can ;e increasing ;. using ,ultiple wireless connections or ;. co,;ining wireless and wired networ8s. - prudent practice is to use ;oth wireless and wired networ8s with the wireless portion ,ini,i2ed to onl. the speci#ic scenarios where deplo.ing a wired connection would ;e cost6prohi;itive. - t.pical e+a,ple is getting a networ8 drop in a ;uilding =either o## the internal 0- or #ro, a telco> and deplo.ing a wireless lin8 #ro, the ;uilding to ca,era locations close to that ;uilding on poles or #ence lines. In an. o# these approaches? C&D'C choice and resolution selection are 8e. #actors in the nu,;er o# ca,eras that can ;e supported. In a wired networ8 where 14 6 144 5;/s networ8s are co,,on? not co,pressing video heavil. can wor8. However? in a wireless networ8? with * 5;/s to 1* 5/;s availa;le total? a single 5I"'G standard de#inition ca,era could consu,e all o# the availa;le ;andwidth ;. itsel#. Si,ilarl.? given the ;andwidth constrains? ,egapi+el ca,eras are especiall. challenges. 'ven with various opti,i2ations? ,egapi+el ca,eras can consu,e #ar greater ;andwidth than standard ca,eras =assu,ing .ou use the sa,e #ra,e rate>. Conclusion
I"Video5ar8et.In#o
*4
3ireless networ8s can solve applications where wired networ8s are #ar too e+pensive. ). relieving the need #or e+pensive construction proEects? video surveillance can ;e deplo.ed in places where it would otherwise ;e cost unEusti#ia;le. However? wireless networ8s o##er #ar greater challenges and ris8s in design and ,aintenance. -s such a clear understanding o# these ele,ents and when to prudentl. use wireless s.ste,s will contri;ute to success wireless video surveillance s.ste,s.
I"Video5ar8et.In#o
*1
Chapter 2:
)PI and S*ste( Integration -utorial
-"Is are the ,ost #reAuentl. ,isunderstood and over6h.ped aspects o# ph.sical securit.. 3hile -"Is can provide great ;ene#its? using the, is ,uch ,ore co,ple+ than o#ten ,entioned in sales calls and ,aga2ines. The goal o# -"Is =or -pplication "rogra,,ing Inter#aces> in ph.sical securit. is to allow di##erent applications to wor8 together. '+a,ples include: 6 Integrating .our DV!/ V! with .our access control s.ste, 6 Integrating .our alar, s.ste, with a central ,onitoring s.ste, 6 Integrating .our I" ca,eras or anal.tics with .our V! 6 )uilding a "SI5 s.ste, that integrates with all .our securit. s.ste,s
Fou ,ost co,,onl. hear -"Is discussed in pre6sales situations where a custo,er or integrator as8s a vendor: BDoes .our s.ste, wor8 with <O<9B where O could ;e an. nu,;er o# securit. s.ste,s ;. an. nu,;er o# ,anu#acturers. The routine answer ;. the sales person is: BSure? we have an -"I.B @or as long as I have ;een in securit. I have ;een hearing this response.
This is the ,ost dangerous and ,isleading state,ent in all o# ph.sical securit.. )ecause it is so co,,on and so dangerous? it is a great place to start reviewing I"Video5ar8et.In#o *2
-"I and S.ste, Integration Tutorial
-"Is.
(esson 345 No suc% t%in" as 6an6 API There is no such thing as <an< -"I. u,erous -"Is e+ist. In larger s.ste,s? hundreds o# -"Is e+ist. Generall.? there is an -"I #or each #unction in a s.ste,. 3ant to watch live video? use the live video -"I. 3ant to change the ti,e? use the ti,e change -"I. 3ant to increase the #ra,e rate #or recording? use the recording #ra,e rate -"I? etc.
(esson 375 Not all &unctions %a'e an API Here<s the #irst gotcha. ot all #unctions have an -"I availa;le. 0et<s sa. .ou need to get a list o# all health alerts #ro, another application. This application ,a. have <an -"I< ;ut not a speci#ic -"I #or sending health alerts. -s .ou can i,agine ;ecause ,ost s.ste,s toda. have hundreds o# #unctions? it is co,,on that do2ens o# these #unctions are not accessi;le via an -"I.
(esson 385 +a'in" an API $oes not mean it will wor* wit% your system 0et<s sa. .ou have Genetec #or .our V! and So#tware House #or .our access control. )oth o# these co,panies certainl. have -"Is ;ut there is no guarantee that these two products will wor8 together. )oth co,panies having -"Is is a pre6 reAuisite #or integration ;ut it is not su##icient. -t least? ;oth o# these co,panies need to wor8 together to ensure the integration wor8s relia;l.. 5an. co,panies certi#. their -"I wor8s with partners ;ut #reAuentl. .our product co,;ination will not ;e included.
I"Video5ar8et.In#o
*%
(esson 395 Doin" t%e Inte"ration a*es ime Vendors o#ten clai, a #ew wee8s #or integration. This can happen ;ut o#ten technical details need to ;e wor8ed out that can ta8e signi#icantl. longer. )e care#ul in the ti,e and dollar a,ount .ou co,,it #or such proEects. This is the t.pe o# ris8 that is o#ten un8nown and un8nowna;le until .ou dig into the technical details a;out how each vendor i,ple,ents their -"Is. Generall.? these proEects are ulti,atel. success#ul? ;ut the ti,e and cost can var..
(esson 3:5 API C%an"es can Brea* ;ou Iust li8e a product? over ti,e? -"Is change. The di##erence is with -"Is? their change can ;rea8 .our s.ste,. !easons #or change include eli,inating ;ugs? enhancing per#or,ance? adding in new #unctionalities. &ther s.ste, depends on those -"Is. 0et<s sa. .our s.ste, wor8s with BVendor )B version %.1. ow let<s sa. BVendor )B co,es out with %.2 ;ut this version B;rea8s the -"IB. In other words? the new version is not ;ac8wards co,pati;le with the old version. Four s.ste, could suddenl. stop wor8ing with BVendor )B i# .ou upgrade Vendor ) to version %.2. The result is .our securit. co,,and center no longer displa.s video or access or whatever the s.ste, that Eust got the upgrade.
(esson 3<5 ;our Stuc* wit% w%at t%e API $oes :nless .our a ver. large custo,er? .ou are stuc8 with whatever the -"I does in whatever wa. it does it. &#ten? #or what .ou need? this wor8s out #ine. However? i# .ou need so,e change #or .our speci#ic use case? this can ;e hard to acco,plish. 5a8e sure so,eone on .our technical tea, 8nows speci#icall. what the -"I can and cannot do so .ou can anticipate an. potential pro;le,s up #ront.
I"Video5ar8et.In#o
*(
I# a change needs to ;e ,ade? the change will usuall. ta8e a lot o# ti,e and testing. This occurs not ;ecause people are slow ;ut ;ecause the vendor ,ust ensure that the. do not ;rea8 the 1444s o# other securit. organi2ations using this -"I. The use o# -"Is are certainl. ;ene#icial #or ph.sical securit. and their use will undou;ted a;l. grow. :nderstanding the realities o# using -"Is will ulti,atel. help us ,a+i,i2e our value o# s.ste, integration.
I"Video5ar8et.In#o
**
How to Integrate Video 3ith &ther S.ste,s
Chapter 13:
How to Integrate Video 1ith "ther S*ste(s
It<s tough and getting tougher to #igure out the ;est approach to integrate video surveillance with other securit. s.ste,s. 3hile the industr. conversation centers on the value o# integration? the real challenge is how to ,a8e this happen? e##ectivel.? cost6e##icientl. and si,pl.. This challenge is growing and is not si,pl. the standard issues in technolog. selection and design. - #ew .ears ago? the options were #airl. clear =i# e+ceedingl. li,ited>. &r spea8ing ,ore precisel.? the option was #airl. clear: The access control s.ste, #unctioned as the co,,and center and the other s.ste,s? such as video #eed into the access control<s plat#or,. Toda.? we have three categories? contenders i# .ou will? #or the role o# ,aster application in securit. s.ste,s: 1. The -ccess Control S.ste,: the classic approach 2. The "SI5 s.ste,: the e,erging trend o# deplo.ing a dedicated application ,anaging traditional securit. s.ste,s %. The Video Surveillance S.ste,: a growing ,ove,ent ;. video vendors to ,anage other s.ste,s 3hich one do .ou choose9 3hich one is ;est9 3hich one will win9 Access Control -ccess control is the ,ost well developed o# the options availa;le? having ;een #ostered over the last decade. 5ost access control s.ste,s can inter#ace with a variet. o# video ,anage,ent s.ste,s. 7e. advantages include the #act that al,ost ever.one has access control and adding in the inter#aces is #airl. ine+pensive. The ,ain custo,er draw;ac8 o# access control s.ste,s as the central plat#or, is that the. tend to li,it %rd part. support to products that ,ost I"Video5ar8et.In#o */
help their i,,ediate sales. The larges incu,;ents such as G'? T.co =So#tware House> and Hone.well have all ;een cited on these issues. -lso? access control s.ste,s al,ost never support other access control s.ste, so i# .ou need to support ,ultiple access control s.ste,s? this generall. will not wor8. PSIM System 3hile "SI5 stands #or the concept o# ,anaging ph.sical securit. in#or,ation? it also covers a group o# co,panies that are ;uilding dedicated applications whose sole purpose is to ,anage securit. s.ste,s such as access control and video ,anage,ent s.ste,s. ota;le vendors include &rsus? "ro+i,e+ and Vids.s. )ecause the. are not owned or controlled ;. access control or video vendors? the. can and do o##er a wide variet. o# support #or di##erent ,anu#acturers. The. also are opti,i2ing their solution #or large6scale securit. ,anage,ent rather than e+tending an e+isting access control s.ste,. The downside is that .ou have to ;u. a new product that is neither cheap nor trivial to i,ple,ent =L144?444 :SD 6 L1?444?444M :SD>. Vi$eo Mana"ement 5ore and ,ore? video ,anage,ent vendors are adding in "SI5 #unctionalities into their s.ste,. @or instance? Video 'OT? traditionall. a video ,anage,ent vendor? is now ,ar8eting a video M "SI5 solution. Verint<s e+tiva and &nSSI<s &cularis are ;ringing in "SI5 #eatures such as ,apping? third part. s.ste, integration? wor8#low ,anage,ent? etc. - 8e. advantage is that it can ;e cheap and eas. to add #unctionalities into a :ser Inter#ace that a custo,er ,a. alread. ;e using. However? li,ited or no support o# other video s.ste,s is an i,portant downside. To ,a8e it even ,ore con#using? two o# the "SI5 vendors? &rsus and "ro+i,e+? o##er power#ul video ,onitoring solutions that provide ;etter large scale ca,era ,onitoring than ,an. video ,anage,ent vendors. Recommen$ations 5a8ing this decision is not eas. as no single approach is ;roadl. applica;le. Fou I"Video5ar8et.In#o *1
should start ;. investigating the a;ilities o# .our current access control s.ste,. This is pro;a;l. the least costl. and si,plest wa. to do integration. I# .ou have concerns with this approaches =and .ou certainl. ,a. ;ecause it has li,itations>? I would then reco,,end investigating the "SI5 providers. This will ;e e+pensive and co,ple+ ;ut the pro;a;ilit. #or integrating all o# .our s.ste,s is high.
I"Video5ar8et.In#o
*8
How to 5igrate #ro, -nalog to I" Ca,eras
Chapter 11:
How to 4igrate ro( )nalog to IP Ca(eras
5igrating #ro, analog to I" can ;e tric8.? ,ainl. ;ecause ,ost ever.one has e+isting in#rastructure in place. Fou rarel. can si,pl. throw out that in#rastructure and start anew 6 the econo,ics usuall. do not support it. )ecause o# that? .ou need to #igure out what to 8eep? what to replace and what to ,odi#.. The issues involved are too co,ple+ to provide a si,ple ;oilerplate .es or no. This report e+a,ines the ,ost critical ele,ents in ,a8ing the transition #ro, analog ca,eras to I" ca,eras so that .ou can ;etter appreciate the issues involved #or .our circu,stances. onetheless? .ou will have to spend signi#icant ti,e learning and evaluating as the issues involved are signi#icant. Here is a su,,ar. o# those 8e. ele,ents:

Deter,ine i# .our DV! supports I" ca,eras Deter,ine what I" ca,era ,anu#acturers .our DV! support I# needed? assess options #or V!s or I" Video 5anage,ent So#tware Deter,ine i# I" ca,eras can eli,inate long distance analog ca;ling Deter,ine i# higher resolution ca,eras can help .ou -ssess the increased ;andwidth i,pact on .our networ8s Deter,ine i# .ou can a##ord increased storage #or ,egapi+el ca,eras
DVR Supports IP Cameras @irst chec8 whether .our DV! supports I" ca,eras. 5ost DV!s that cost ,ore than L%?444 :SD usuall. supports so,e #or, o# I" ca,eras toda.. However? ,ost o# the ,ore <;udget< t.pe DV!s do not. Fou should deter,ine this #irst ;ecause it is the 8e. ele,ent in deter,ining how co,ple+ adding in I" ca,eras will ;e. I# .our DV! does not support I" ca,eras?
I"Video5ar8et.In#o
*$
.ou have a #ew options? none o# which I thin8 are ver. attractive: =1> .ou could ,onitor the I" ca,eras directl. with no recorder? =2> .ou could set up a separate V! to record the I" ca,eras or =%> .ou could decode the I" ca,era<s video strea, to record the, on .our e+isting DV!s. 5ost pro#essional securit. organi2ations want a single video ,anage,ent s.ste, to record and access all ca,eras which ,eans that .ou either wor8 with what .ou have or replace it. -%at IP Cameras ;our DVR Supports I# .ou DV! supports I" ca,eras? .ou de#initel. need to #ind out what ,anu#acturers and ,odels o# I" ca,eras the. support. 5an. DV! suppliers onl. support 1 or a s,all nu,;er o# I" ca,era ,anu#acturers. This can ;e reall. con#using and surprising co,ing #ro, the analog ca,era world. 3ith analog ca,eras? no one worried a;out whether a DV! could support a #i+ed ca,era ;ecause once .ou supported 1 analog ca,era? .ou supported the, all. However? with I" ca,eras? .ou have to chec8 ever. ti,e #or not onl. ,anu#acturer support ;ut #or speci#ic ,odel support =i.e.? a DV! ,anu#acturer ,a. support the -+is 241 ;ut not the -+is 221>. Deter,ining what I" ca,eras a DV! supports is ver. i,portant ;ecause di##erent ,anu#acturers speciali2e in di##erent t.pes o# products. I# .our DV! onl. supports 1 or 2 ca,era ,anu#acturers? this could cause signi#icant pro;le,s. @or instance? there are specialists in high end? standard de#inition ca,eras =-+is>P ;udget standard de#inition ca,eras =-CTi>P ine+pensive ,ulti6,egapi+el ca,eras =-recont Vision>P high end ,ulti6,egapi+el ca,eras =IQinvision>? etc. Fou need to deter,ine what t.pes o# I" ca,eras .ou need and whether those are supported ;. .our DV!. These #irst two points will help .ou understand the degree o# di##icult. o# adding in I" ca,eras. NVRs or IP Vi$eo Mana"ement Software -t this stage ,an. will reach a point where .ou need to consider replacing .our I"Video5ar8et.In#o /4
DV! s.ste,. The e,erging alternative are designed to support do2ens o# I" ca,eras. I# .ou get to this point? this will ;e a challenge in and o# itsel#. There are do2ens o# co,panies that o##er V!s or I" Video 5anage,ent so#tware. @urther,ore? i# .ou head in this direction .ou will need to deter,ine how to support .our e+isting analog ca,eras. )ecause I" Video 5anage,ent So#tware onl. supports I" video strea,s? .ou will need to purchase encoders to convert the analog video strea, #ro, .our ca,era into an I" video strea, that the I" Video 5anage,ent so#tware can handle. 'ncoders are #airl. e+pensive =L%44 6 L/44 :SD per ca,era> so it ,a. ;e worthwhile ;ut it is not without its costs. This covers the #unda,ental product options and choices. To deter,ine i# the ,igration is worth it? #ocus on the ne+t two ite,s. Eliminate (on" Distance Analo" Ca#lin" -ll ca,eras need to ;e connected to a video recorder. How the. are connected can var. greatl.. The ,ost co,,on ,eans #or analog ca,eras is to use a dedicated coa+ial ca;le to connect the ca,era to the DV!. Indoors and over short distances? this is usuall. Auite si,ple to do. However? i# .ou need to go long distances? outdoors or through areas where it is hard to run a new dedicated ca;le? analog ca,eras can ;eco,e pro;le,atic. I# .ou have ,ultiple ;uildings or outdoor areas to protect? .ou ,a. not ;e currentl. using surveillance ca,eras or i# .ou are .ou had to resort to e+pensive proprietar. trans,ission s.ste,s. This is the ,ost valua;le and power#ul use o# I" ca,ears. 3ith I" ca,eras? .ou have the potential o# reusing e+isting networ8s in .our #acilities. Fou also can use low cost I" wireless eAuip,ent to add ca,eras in distant or outdoor locations. To the e+tent that this situation applies to .ou? .our ,otivation to ,ove to I" ca,eras should ;e stronger. It can either reduce costs ;. thousands o# dollars co,pared to e+isting i,ple,entation or ena;le .ou to add new ca,eras in places that would have ;een previousl. cost prohi;itive. I"Video5ar8et.In#o /1
=se of +i"%er Resolution Ima"es I" ca,eras o##er the potential to capture and record ,uch higher resolution i,ages than analog ca,eras. 3hile the ,a+i,u, resolution o# ,ost I" ca,eras is the sa,e as ,ost analog ca,eras? one t.pe o# I" ca,era? the ,egapi+el ca,era? can o##er #ar greater resolution. Fou should deter,ine how and where .ou can ,a8e ,ost use out o# ,egapi+el ca,eras. 7e. deter,inants are =1> the greater the area .ou want to cover and =2>the higher .our need to see details. @or e+a,ple? a par8ing lot or cashier<s station. ). contrast? i# .ou are o;serving a s,all o##ice roo, and Eust need to 8now when so,eone was inside? a traditional standard de#inition analog ca,era will do #ine. 5egapi+el ca,eras co,e with two huge i,pacts that .ou ,ust consider when ,igrating #ro, analog ca,eras: ;andwidth and storage. Assess t%e Ban$wi$t% Impact 3hen ,igrating #ro, analog to I"? i# .ou 8eep the resolution .ou record at the sa,e? the i,pact on ;andwidth =.our co,puter networ8> should ;e ,ini,al. @or instance? ,ost co,,ercial users record at * #ra,es per second at CI@ =%24 + 2(4 pi+els>. -t these levels? ;andwidth consu,ption is Auite low =under .* 5;/s> relative to toda.<s networ8s =144 5;/s MM>. 'ven with a #ew do2en ca,eras? this should not ,a8e a signi#icant i,pact on even lower end switches. However? i# .ou want high resolution or #ra,erates? then .ou need to start care#ull. assessing the i,pact. 3ith these conditions? each ca,era can consu,e *5;/s to (* 5;/s? which starts adding up. 3hile .ou can purchases networ8ing eAuip,ent that can handle 14445;/s or ,ore? .ou should not assu,e that this is alread. in place and that .ou can Eust plug this in. Fou certainl. should test the ;andwidth load ;e#ore deplo.,ent. Fou ,a. need to consider one o# the #ollowing two options:
I"Video5ar8et.In#o
/2
1. :se a separate I" networ8 #or the ca,eras. 2. :pgrade .our e+isting networ8ing eAuip,ent to ,a8e sure that it can support the load. )oth are certainl. e+pensive and can have a signi#icant operational and political i,pact with .our IT<s organi2ation. Though this can ;e acco,plished? do not ta8e it #or granted as the cost and co,ple+it. can ;e signi#icant. Assess t%e Stora"e Impact In a si,ilar ,anner? increasing the video Aualit.? certainl. i,pacts storage needs. I# .ou use DV!s? .ou are li8el. used to ;u.ing storage ;undled with the DV! =e.g.? a DV! with 2*4 G) or *44G)s o# storage #or 1/ ca,eras>. 3ith I" ca,eras and? especiall. with ,egapi+el? .ou can easil. ;e loo8ing at 1T) per ca,era? which is a ver. signi#icant increase. This could increase the cost o# .our s.ste, ;. tens o# thousands o# dollars. Fou will need to ;etter deter,ine how signi#icant this will ;e and .our willingness to spend ,ore #or storage. So,e organi2ations will #ind it to ;e no ;ig deal ;ut others ,a. ;e shoc8ed. Conclusion Hope#ull. this helps identi#ies 8e. points so .ou can ;etter assess .our situation. "lease as8 Auestions? add other points and de;ate the appropriateness o# the reco,,endations ,ade.
I"Video5ar8et.In#o
/%
Director. o# &n60ine Video Surveillance Tutorials
Chapter 12: -utorials
Director* o "n5,ine Video Surveillance
)asic 0ighting #or Ca,eras #ro, a+is Description: B0ighting is a 8e. ele,ent in providing high Aualit. i,ages. 0ighting is co,,onl. ,easured in lu+. 5easuring lu+ is not eas. #or ,ost people. This resource #ro, -+is provides a slideshow o# e+a,ple i,ages with lu+ ratings. I# .ou 8now what lu+ is? this will give .ou a ;etter appreciation o# how to guess / ,easure how ,uch lu+ in an. scene.B )asic &verview o# Securit. Ca,eras #ro, electronichouse Description: BI# .ou are not #a,iliar with securit. ca,eras at all? this article will provide a nice introduction. Though the article is #ocused at the ho,e surveillance ,ar8et? the in#or,ation provided is generall. applica;le to all seg,ents o# video surveillance.B
Introduction to 0enses #ro, cctv#ocus Description: B0enses are a critical ;ut o#ten overloo8ed co,ponent o# video surveillance. In this tutorial ;. Vlado Da,Eanovs8i? the author o# the BCCTV )i;leB? Vlado e+plains the #unda,entals ele,ents o# lens design and selection. Fou should have a ,oderate understanding o# ca,eras to get the ,ost out o# this tutorial.B IP Cameras How to Set up -n I" Ca,era #ro, a+is Description: BThis is a ver. nice video #ro, -+is on how to set up/con#igure an I" ca,era. -s long as .ou 8now the ;asics o# I" ca,eras? this tutorial should ;e ver. I"Video5ar8et.In#o /(
use#ul. It provides a ver. clear sense o# si,ple and power#ul con#igurations that can ;e done to opti,i2e the use o# I" ca,eras.B -n Introduction to H.2/( #ro, securit.in#owatch Description: B- ver. co,prehensive introduction to H.2/( that covers ;oth the ;ene#its and the technolog. details o# H.2/(. -s this is a Auite technical article? .ou should ;e co,#orta;le understanding other C&D'Cs such as 5I"'G and 5"'G6( ;e#ore reading this article.B 3hat is an I" Ca,era9 #ro, a+is Description: B@or the ver. ;asics in what an I" ca,eras is? start with this tutorial #ro, -+is. @ro, this tutorial? .ou can watch a video on I" ca,eras or e+plore details on #urther pages inside. This tutorial does not reAuire technical e+pertise and is good #or all audiences.B Introduction to "ower over 'thernet #or I" Ca,eras #ro, sd,,ag Description: B:sing "ower over 'thernet with I" ca,eras is a ,aEor change #ro, traditional analog ca,era deplo.,ent. It o##ers signi#icant reduction in co,ple+it. and cost. evertheless? it is a new technolog. to ,an. in video surveillance/ph.sical securit.. this tutorial does a nice Eo; o# e+plaining the issues involved. 5oderate understanding o# co,puters and networ8s will ;e help#ul #or this one.B Me"api0el Cameras Introduction to 5egapi+el Ca,eras #ro, securit.in#owatch Description: BThis tutorial provides an introduction to the use and ;ene#its o# ,egapi+el ca,eras. 3hile a ;asic ;ac8ground in video surveillance is help#ul? .ou do not need to 8now ,uch a;out ,egapi+el ca,eras to learn #ro, this piece.B
I"Video5ar8et.In#o
/*
NVRs>IP Vi$eo Software Introduction to Video 'ncoders #ro, sd,,ag Description: B'ncoders are an i,portant co,ponent o# networ86;ased video ,anage,ent and recording. 'ncoders allow analog ca,eras to wor8 with V!s. This article is an e+cellent overview o# what encoders are and what role the. pla. in video surveillance s.ste,s. It does not reAuire special technical e+pertise.B
Vi$eo Analytics 3e;inar Introduction to Video -nal.tics #ro, isc%/* Description: BThis tutorial is a we;inar where .ou can watch and listen as e+a,ples o# various video anal.tics are de,onstrated. This is good #or the ;eginner as it does not reAuire an. speci#ic technical e+pertise or ;ac8ground.B How to "osition Ca,eras #or "eri,eter Violation #ro, ioi,age Description: B&ne o# the ,ost #unda,ental ;ut underappreciated co,ponents o# using video anal.tics is setting up ca,eras. This tutorial? #ro, IoI,age? does a good Eo; o# honestl. e+a,ining the ,an. details i,portant to ,a8ing peri,eter violation wor8 in the #ield. 'ven i# .ou do not 8now ,uch a;out ca,era #unctionalities? viewing the tutorial should give .ou a #eel o# the steps needed.B How to Test "eri,eter Violation -nal.tics #ro, ioi,age Description: B5ar8eting videos o#ten leave the i,pression that a video anal.tics s.ste, wor8s i# it can si,pl. alert when a person is casuall. wal8ing through the scene during the ,iddle o# the da.. This tutorial? #ro, IoI,age? de,onstrates the ,an. di##erent scenarios that a production video anal.tic should ;e a;le to handle. Fou ,a. not actuall. ever do this test ;ut reading this tutorial should give .ou a ;etter appreciation #or the real ,easures an adversar. ,a. atte,pt to ;eat .our I"Video5ar8et.In#o //
s.ste,. This tutorial does not reAuire an. technical e+pertise.B
Networ*in" Designing I" etwor8s #or Video Surveillance #ro, isc%/* Description: BThis is a ,id6level to advanced tutorial that e+plains the 8e. points and design ele,ents in designing I" networ8s #or Video Surveillance. This tutorial is a 1 hour we;inar hosted ;. Cisco S.ste,s. The #irst part will ;e valua;le #or a general audience as it reviews the ;asics o# networ8ing. The second part on QoS techniAues reall. reAuires a ;ac8ground in networ8ing to ;e use#ul.B Introduction to I" 5ulticasting #ro, indigovision Description: B5ulticasting can provide signi#icant ;andwidth reduction when ,an. users want to view the sa,e live strea,. However? ,ulticasting is not si,ple. This tutorial? #ro, IndigoVision? provides a nice overview o# the ;ene#its? challenges and steps needed in deciding and using ,ulticast. The tutorial assu,es a ,oderate level o# IT/networ8ing concepts.B Technical )asics o# I" etwor8s #ro, a+is Description: BI# .ou are not co,#orta;le with ;asic technical ter,s and concepts o# I" networ8s ;ut want to learn the,? this is the right tutorial #or .ou. It is a gentle overview o# the ;asic ele,ents i,portant to .ou in dealing with I" networ8s #or video surveillance.B
Con'er"ence -n Introduction to "SI5 #ro, securit.drea,er I"Video5ar8et.In#o /1
Description: B"SI5? or "h.sical Securit. In#or,ation 5anage,ent is an i,portant e,erging #orce within securit. and video surveillance. 3ith ca,era counts continuing to grow and video anal.tics generating in#or,ation? the de,and #or an organi2ed and s.ste,atic wa. to ,anage this in#or,ation
I"Video5ar8et.In#o
/8
II
Examining Key Trends and Technologies
Chapter 1!:
I- Is Not -a6ing "ver Securit*
So ,uch tal8 toda. #ocuses on the power o# IT and what IT is doing to securit.. 3hile securit. ,anagers will certainl. leverage that technolog.? IT is not replacing or ta8ing over securit..
-uto,ation is a power#ul econo,ic #orce? one that will ulti,atel. ,a8e IT irrelevant to ph.sical securit.. It ,a. see, parado+ical ;ut the sa,e #orce that ,a8es in#or,ation technolog. u;iAuitous will ,a8e IT irrelevant to ph.sical securit.. "h.sical securit. will certainl. use ,ore technolog. than ever ;e#ore ;ut the technolog. will ;eco,e easier to use and deplo.. -s it ;eco,es easier to use and deplo.? the need #or IT decision ,a8ing and IT personnel will di,inish and ;eco,e a ,inor #actor. Co,panies li8e IoI,age are alread. showing us this #uture while ;ig IT co,panies such as Cisco and I)5 are stuc8 in the past. In the securit. industr.? it<s eas. to #ear that we are ;eing go;;led up ;. IT ;ut it<s Eust a phase.
%e ROI of Automation an$ Simplicity This is eas. to predict ;ecause the econo,ics de,and this and the histor. o# other #ields de,onstrate this. IT has ;een assi,ilated into ever. depart,ent in the enterprise and a#ter the initial introduction? control alwa.s returned to the depart,ent. The #irst #inancial s.ste,s and C!5 s.ste,s were huge co,ple+ proEects that de,anded custo, so#tware develop,ent and e+tensive on6site ad,inistration. IT o;viousl. had to ;e heavil. involved. Toda.? the sales ,anager can get up and running #or little ,one. and hassle #ro, Sales#orce.co,? etc. The sales ,anager has a huge incentive to si,pli#. ;ecause he does not want his sales operations hindered ;. never6ending IT proEects. He will engage IT #or support and services ;ut he uses their input to ,a8e his own #inal decision. The sa,e will happen #or securit.. 3henever technolog. is co,ple+ and unpredicta;le it reAuires e+tensive anal.sis?
planning and #ield integration/ support. This is a characteristic o# earl. stage technolog.. The #irst DV!s had this characteristic. The #irst I" ca,eras? ,egapi+el ca,eras? anal.tics? etc did as well. 14 .ears ago DV!s were a ,aEor IT proEect and now the. are u;iAuitous? can ;e purchased at Costco and installed ;. the store ,anager. Technolog. vendors saw that their sales were li,ited due to the e+tensive on6site integration? testing and planning needed. To increase their own sales? the. wor8ed hard to si,pli#. and eli,inate deplo.,ent challenges. -s this happened? IT ;eca,e less and less i,portant and the DV! needed onl. ru;;er sta,p approval #ro, ,ost IT depart,ents. The decision ,a8ing returned to the securit. ,anager who now onl. #ocused on which product ;est helped hi, ,eet his securit. goals.
Cisco an$ IBM $o not "et I In perhaps the greatest iron. in our industr.? the two ;iggest <IT< co,panies Eust do not get it. The. ;ring to ,ar8et incredi;l. co,ple+? e+pensive s.ste,s that de,and e+tensive #ield integration. @ro, Cisco and I)5<s perspective? ,a.;e this loo8s li8e great ;usiness ;ecause the products are e+pensive and the #ollow6on services are su;stantial. )ut #ro, the custo,er<s perspective? this is aw#ul. This is si,pl. a ta+ on the custo,er? dropping the !&I and ,a8ing the ;usiness case ,ore di##icult. 3ith Cisco and I)5? .ou have to initiate a huge proEect? get the CI& involved? spend ,onths planning and deplo. a tea, o# engineers/consultants. IoI,age is showing us the #uture o# advanced IT technologies in the ph.sical securit. space. 0et<s contrast what IoI,age is doing to what Cisco and I)5 have done.
IoIma"e $oes "et I IoI,age<s slogan is BDesigned #or Si,plicit.B and .ou can see #ro, their product design to their distri;ution strateg.? that ,a8ing it si,ple to purchase? setup and deplo. is a 8e. ;usiness o;Eective. Do .ou 8now where .ou can ;u. IoI,age products #ro,9 orthern video and now? SuperCircuits. That<s right? SuperCircuits 66 The ,aga2ine .ou get with the L12.$* pinhole sp. ca,eras now o##ers IoI,age video anal.tic ca,eras. This is a great thing #or securit. ;u.ers ;ut a signal o# the pro;le,s #or Cisco and I)5.
3hile Cisco and I)5 are ,a8ing it e+pensive and co,ple+ to use advanced technologies in ph.sical securit.? IoI,age is ,a8ing it eas.. IoI,age is ,a8ing it so eas. that it<s reducing the need #or IT to ;e involved. IoI,age let<s the securit. ,anager concentrate on ,eeting his securit. goals. So while Cisco and I)5 ,a. ;e #riends with the CI&? co,panies li8e IoI,age will ena;le the highest !&I ;. delivering a##orda;le and eas. to deplo. technologies. IoI,age<s strateg. shows the #uture o# ph.sical securit.. It<s a #uture o# u;iAuitous technolog. #reed #ro, the cost and co,ple+it. o# ;ig IT. Though technolog. has ;een and will continue to ;e a 8e. #orce in ph.sical securit.? IT<s i,portance is Eust a phase. IT does not get this ;ut Securit. 5anagers and Integrators should have #aith in this.
Chapter 1$:
1ill Securit* Integrators Survive7
-l,ost all securit. ,anagers use securit. integrators. -s such the #ate o# securit. integrators and the value o# continuing to use securit. integrators is a 8e. Auestion toda.. 5an. ;elieve securit. integrators are deadP wal8ing dinosaurs who are o;livious to their i,pending e+tinction. Indeed? ,an. new IT entrants certainl. wish that securit. integrators =and ph.sical securit. ,anagers> are wiped out. Despite this ;elie# and hope? is this reall. the case9 Are security inte"rators $estine$ to fail, o? I #elie'e security inte"rators? as a w%ole? will sur'i'e. I ;elieve the detractors have ,ade % ,ain ,ista8es:
Detractors loo8 at convergence as a recent pheno,enon whereas securit. integrators have ;een adapting #or the last 14 .ears Detractors do not appreciate the s8ills that securit. integrators possess Detractors view IT as a disruptive innovation when it is trul. a sustaining one
A$aptin" for 4@ ;ears Securit. integrators have ;een assi,ilating IT s8ills #or the last 14 .ears. 3hile a lot o# an+iet. e+ists over I" ca,eras and V!s? the technical challenges were #ar worse 8 to 1* .ears ago. -t that point? integrators were deplo.ing their #irst DV!s or networ8 ;ased -ccess Control s.ste,s. 5ost had no IT s8ills. 5an. did not even 8now what an I" address is. &ver the .ears? with education and on the Eo; e+perience? the situation has changed dra,aticall.. Toda.? ,ost securit. technicians have at least ;asic IT s8ills and ,an. are #airl. sophisticated. I" ca,eras and V!s present new technical challenges ;ut the. are e+tensions o# the ;asic s8ills securit. technicians have ;een learning #or .ears. I a, not contending that securit. technicians are as strong in IT as IT technicians. However? there are ( ver. real aspects that a##ect the co,petitiveness ;etween IT and securit. integrators:

IT technicians are ,uch ,ore e+pensive than securit. technicians Securit. technicians have a good ;asis #or the tas8s needed #or I" securit. s.ste,s
-s I" ;ased securit. s.ste,s ,ature? the. are ;eco,ing easier #or non6IT e+perts to use - lot o# what IT technicians 8now is over8ill #or I" securit. s.ste,s
)ecause the. have ;een adapting #or the past 14 .ears? securit. integrators can o##er ,an. o# the IT s8ills needed at less cost than IT integrators. This is an under6appreciated #actor in wh. securit. integrators will survive.
Security Inte"rator6s S*ills 5an. underesti,ate the i,portance o# securit. integrator<s s8ills and the value those s8ills will continue to have in proEects. Two 8e. issues e+ist:

Securit. Integrators have ,an. 8e. s8ills that IT Integrators lac8 'ven in I" securit. s.ste,s? ,ost o# the integration wor8 is not IT
Good securit. s.ste,s integration reAuires e+tensive design and i,ple,entation precision. It is #ar ,ore ,ore involved than si,pl. installation plus IT tas8s. Securit. integrators have ;een learning s8ills #or .ears that IT integrators totall. lac8. Good integrators in an. #ield participate in design. Iust #or video surveillance? securit. s.ste,s integrators ,ust ;e a;le to:
-nticipate what areas and assets reAuire protection =,an. end users need guidance here> Deter,ine how to protect those assets with what ca,eras and what positioning :nderstand the li,itations o# the products availa;le #or protecting/i,aging :nderstand the environ,ental li,itations and how to acco,,odate the, Deter,ine and eli,inate gaps in coverage
one o# these are IT tas8s ;ut all o# these are essential in integrating a high Aualit. video surveillance s.ste,. These s8ills are needed throughout the i,ple,entation and are not a distinct part o# a;stract design. !arel. can securit. eAuip,ent si,pl. ;e installed. It reAuires s8ill and Eudg,ent in how to adapt to on6site issues:
Iudg,ent and s8ill in #inal ca,era positioning is critical in ensuring the right shot Ca,era settings and lenses will need to ;e adEusted to opti,i2e i,age Aualit. Ca,eras ,a. need to ;e ,oved due to an un#oreseen i,ple,entation issue &n site ,anagers ,a. o;Eect to aesthetics and integrator will need to #ind new positioning
-gain? none o# these are IT tas8s and none o# these go awa. with I" ;ased s.ste,s. Fou need to ,aster these aspects #or good securit. integration. The average securit. integrator has this. The average IT integrator does not. The sa,e pattern e+ists in developing policies and ;est practices #or using securit. s.ste,s. Securit. Directors routinel. e+pect and lean on their integrators to help teach the, and share ideas on how to ;est use the technolog. #or securit. o;Eectives. IT s8ills are o# little help here unless .ou 8now the application and the issues involved in ph.sical securit.. 84K o# the wor8 involved in securit. s.ste,s integration is in the areas I have Eust outlined. The IT side is certainl. valua;le ;ut as a ,atter o# ti,e and e##ort? it is a rather s,all portion o# overall IT proEects. -s such? it is a natural candidate #or securit. integrators to si,pl. e+pand and integrate into their services. -nd as I have discussed a;ove? this is is part o# a long ter, trend that securit. integrators have ;een doing #or .ears.
I is a Sustainin" &orce to Security Inte"rators The e,phasis on IT ;eing a <disruptive< technolog. to securit. is ,isleading. 5an. thin8 disruption is a #actor o# how sophisticated or power#ul a technolog. is. In a ;usiness conte+t? that does not ,a8e a technolog. disruptive. Technologies onl. disrupt ;usinesses when the. disrupt ;usiness ,odels. The widel. held theor. o# innovation contends that i# a new technolog. ena;les incu,;ents to ,a8e ,ore ,one. #ro, their ;est custo,ers in the sa,e wa. the. have historicall.? the incu,;ents usuall. win. Securit. integrators can ,a8e ,ore ,one. #ro, their ;est custo,ers ;. selling I" ;ased securit. s.ste,s. -s such? innovation theor. holds that securit. integrators should survive. Iust li8e the. did ;e#ore? securit. integrators are still selling products? integration services and ,aintenance services. "lus? the revenues per deal have generall. increased. In the o#ten cited scenarios where incu,;ents where 8illed? it was ;ecause prices were
radicall. di##erent =e.g.? ,ini6co,puters vs "Cs> or the ;usiness ,odel switched #ro, selling products to su;scriptions =e.g.? SaaS>. This is Eust not the case here. There<s no reason to thin8 securit. integrators will retreat and growing evidence that the. are responding. -ll the ;ig securit. integrators are #inanciall. ,otivated to co,pete and the. have resources the. can invest in IT. Iust li8e ,an. other industries? securit. integrators will engage in training and will hire new personnel with appropriate s8ill sets? assi,ilating the, into their organi2ation. -nd ;ecause securit. integrators have e+cellent e+isting s8ills in the #unda,entals o# securit. s.ste,s? the. will have a ;ig advantage over IT integrators tr.ing to learn the space? relationships and i,ple,entation details that integrators have ,astered over the .ears.
Conclu$in" %ou"%ts !unning a securit. integrator? I have lived through all o# these ele,ents #irst hand. -t that ti,e? I was the IT outsider ;rought in to help the transition. However? it was I who assi,ilated ;ecause that ,ade the ;est ;usiness sense. &# course? I ;rought in new training? practices and s8ills that helped grow the ;usiness. evertheless? we used the core group o# e+isting securit. technicians as the ;asis? i,proving their IT s8ills and supple,enting the, with a s,all nu,;er o# strong IT engineers. It was si,ultaneousl. less disruptive? ,ore pro#ita;le and allowed us to e+ecute on the ,an. ph.sical securit. related details that the IT engineers would have ta8en a long ti,e and a lot o# ,one. to sort through. Fou ,a. have a couple o# counterargu,ents: Counterar"ument5 My Security Inte"rator is Ba$ It happens. )ut consider that a;out 1/% to 2/% o# all IT proEects #ail. 5a8ing securit. into IT is no panacea. IT has plent. o# its own issues. Counterar"ument5 I is t%e future A It %as to ta*e o'er Security To the e+tent that co,puters are replacing electronics? a;solutel.. Securit. s.ste,s will ;eco,e an IT specialt.? Eust li8e historicall. securit. s.ste,s were a specialt. o# low voltage electronics. However? the co,panies that succeed in securit. as an IT specialt. are li8el. to ;e the traditional securit. integrators who evolve into this role.
Some Security Inte"rators %a'e to fail Certainl.? so,e will #ail. So,e alwa.s #ail ;ut the #ailures will ;e ,ore an issue o# poor individual e+ecution that it is that the whole industr. will collapse. I inte"rators %a'e a lot to offer I agree. 0oo8 to see IT engineers hired into e+isting securit. s.ste,s integrators or see the, start their own specialt. shops dedicated to securit. s.ste,s. I a, onl. o;Eecting to ;ig IT integrators co,ing in and wiping out securit. integrators. There is alwa.s roo, #or new s8ills and new talents to grow an industr..
Chapter 1&:
Should I 8se IP Ca(eras7
I" ca,eras have ;eco,e accepted ;. the securit. industr.. Fet ,ost ca,eras are still analog and ,ost video ,anage,ent s.ste,s are still DV!s. 3hen and how do we ,a8e the transition9 Is it a #ast transition9 3hen does a securit. ,anager? ,anu#acturer or integrator 8now when to ,a8e the ,ove9 Though the ;ig picture see,s settled? with ,uch o# the actual transition still co,e to? how to e+ecute and navigate the transition ;eco,es a critical ;usiness decision.
Bey Strate"ic Points To help ,a8e this transition? here are % 8e. strategic points that shape the ti,ing and e+ecution o# transition tactics.
The larger the #acilit. ;eing secured? the ,ore valua;le an i,,ediate transition to I" ca,eras. The ,ore ,ature ,egapi+el ca,eras ;eco,e? the ,ore valua;le an i,,ediate transition to I" ca,eras. DV!s will continue to catch up to V!s and will as such e+tend the li#e o# analog s.ste,s.
This report e+a,ines these 8e. strategic points and concludes with speci#ic reco,,endations #or integrators and end6users.
Strate"ic Point 345 %e (ar"er t%e &acility The larger the #acilit. ;eing secured? the ,ore valua;le an i,,ediate transition to I" ca,eras. It is not so ,uch how ,an. #acilities ;ut the si2e o# each speci#ic #acilit.. )ecause o# the intrinsic li,itations o# coa+ial ca;le? when #acilities ;eco,e too large? the costs o# s.ste, installation increase dra,aticall.. Thin8 o# o##ice towers? corporate
ca,puses? ,ilitar. ;ases. 0ow cost coa+ial ca;le runs could not solve the pro;le,. "roprietar. networ8s were needed. The eli,ination o# proprietar. networ8s is the one advantage o# I" ca,eras that dwar#s all others and has ;een driving I" ca,eras/encoders. This is where the ;usiness case is a;solutel. roc8 solid. @or large scale surveillance proEects? .ou can save L1?444 to L(?444 per ca,era relative to analog long distance trans,ission s.ste,s. I# .ou can eli,inate trenching? the cost savings are even ,ore dra,atic. It is no surprise that ,ost o# the ;iggest I" ca,era s.ste,s are a,ong schools? corporate ca,puses? ,unicipalities? the ,ilitar.. That<s not to sa. that I" ca,eras are not deplo.ed elsewhere ;ut ,an. i# not ,ost o# the ;iggest success stories are in applications where long distances e+ist ;etween ca,eras. 0i8ewise? we should not ;e surprised that Auic8 serve restaurants? ;an8 ;ranches? s,all and ,ediu, si2e ;usinesses and other organi2ations with s,all #ootprints are slow in the upta8e o# I" ca,eras. Coa+ wor8s Eust #ine there ,a8ing the ;usiness case ,uch harder to Eusti#..
Strate"ic Point 375 %e more mature me"api0el cameras #ecome 'cono,icall. spea8ing? the increase in Aualit. ;etween standard de#inition I" ca,eras and analog ca,eras recorded ;. a DV! is ,ini,al. The Aualit. o# I" ca,eras is certainl. ;etter ;ut it is not so ,uch ;etter that ,an. ,ore cri,es can ;e solved. 3ithout a clear and si2a;le increase in such drivers? the Aualit. o# I" ca,eras does not drive I" adoption =that does not ,ean I" wonRt ;e adopted ;ut it is ,ore li8el. I" is adopted ;ecause o# strategic point S1 and the Aualit. is a nice throw in>. ). contrast? ,egapi+el ca,eras a;solutel. have the potential to solve ,ore cri,es. 3e are seeing the ;eginning o# this with the use o# ,egapi+el ca,eras in casinos. ). ;eing a;le to show a level o# detail i,possi;le with analog ca,eras? losses are ;eing prevented and ,itigated? generating si2a;le ;usiness value to the organi2ation. However? the ;usiness case o# ,egapi+el ca,eras is still wea8 due to its increases in overall s.ste, cost. It is still ver. unclear when and how those costs and co,ple+ities
will ;e overco,e? triggering widespread ,ainstrea, adoption. 3hile ,egapi+el has the potential? it is not .et actuali2ed. This will hasten the transition ;ut when and how9
Strate"ic Point 385 DVRs will continue to catc% up to NVRs &ne o# the ,ost interesting and underappreciated ele,ents in the transition to I" ca,eras is how DV! ,anu#acturers have responded in this transition. This undou;tedl. will continue? ,a8ing it easier to e+tend the li#e o# analog ca,eras. Here are * areas where DV!s have traditionall. ;een #aulted in co,parison to V!s and how DV!s have narrowed the gap: IP camera support: -l,ost all ,ainstrea, DV!s have ;eco,e h.;rid s.ste,s supporting a wide variet. o# I" ca,eras. This trend will continue as the technical i,ple,entation is not ver. hard and custo,ers clearl. want the #le+i;ilit.. 3hile h.;rid DV!s will not support as ,an. ;rands o# ca,eras as V!s? the range o# support is li8el. to ;e good enough #or ,ost users. -nd given? the deep installed ;ase? h.;rid DV!s will o#ten have an econo,ic advantage over s.ste, that reAuire I" ca,eras or encoders. Remote Access: 3hile earl. DV!s ,ight have ;een li,ited in re,ote access? toda. all DV!s o##er a variet. o# wa.s and #unctions #or re,ote access including thic8 client and we; access. @ro, a custo,erRs perspective? the di##erence ;etween DV!s and V!s will rarel. ;e noticea;le. Scala#ility: 3hile V!s had the earl. head start here? it is co,,on #or toda.Rs DV!s to ;e a;le to ,anage s.ste,s o# thousands o# ca,eras. DV!s o##er health ,onitoring? centrali2ed ad,inistration? virtual ,atri+es? etc.? etc. This is not a clai, that DV!s are ;etter or are so,ehow going to 8noc8 V!s out. Si,pl. that DV!s have addressed the 8e. de#iciencies ,a8ing it hard #or I" to win solel. on this point. Inte"ratin" Applications: DV!s have alwa.s ;een strong at integrating with access control? intrusion dection? "&S? -T5s? etc. I #ind clai,s ;. either side on this point to ;e ,ore ,ar8eting h.pe than actual di##erentiation. I suspect ,ost custo,ers will see that either t.pe supports their needs.
Analytics: 3ith the rise o# h.;rid s.ste,s and the continued increase in C": speeds? DV!s are ;eco,ing power#ul anal.tic plat#or,s. The #act that DV!s are h.;rid s.ste,s now ,eans the. can support the sa,e &V or IoI,age ca,eras that an V! can. The #act that lots o# e+tra C": speed can ;e o;tained in DV!s #or ,ini,al cost? ,eans that DV!s are going to ;e running anal.tics inside their s.ste,s. 3ith dual and Auad core ;eco,ing co,,on place? the econo,ics o# per#or,ing anal.tics in DV!s are ;eco,ing ver. co,petitive relative to s,art ca,eras.
So ,an. o# the core I" ca,era advantages have ;een co6opted ;. DV!s. Though it certainl. will not stop I" ca,eras? this is going to ,a8e #urther inroads harder and rein#orce the value o# e+isting and replace,ent analog ca,eras.
Recommen$ations 0etRs start with general reco,,endations that appl. across the industr. and then e+a,ine speci#icall. end6users and integrators. General !eco,,endation S1: %e "rowt% is in lar"e facilities I# .ou are loo8ing to grow responsi;ilities in new areas? the growth area will certain ;e large #acilities. 3h.9 )ecause I" ca,eras change the ;usiness ,odel o# deplo.ing ca,eras in large #acilities and areas. 3here once it was too e+pensive to deplo.? I" is ena;ling new use o# ca,eras. 3e will certainl. see this continue in schools? corporate ca,puses? ,unicipalities? outdoor #acilities? an.where that long distances separate ca,eras #ro, recording/,onitoring stations. General !eco,,endation S2: %e a#solute $ecline in analo" cameras an$ DVRs will #e slow )ecause DV!s are ,oving up and analog ca,eras will re,ain a good value #or s,aller #acilities? e+pect the decline in the use o# analog ca,eras and DV!s to ;e slow. In other words? it is ver. unli8el. that the. we will see a ,ass e+odus #ro, these s.ste, in the ne+t * .ears. This should change as the price co,petitiveness o# I" ca,eras increases and as V! solutions ;eco,e si,pler to setup and ,anage. However? this is a process that will evolve over a nu,;er o# .ears.
General !eco,,endation S%: Pay Close Attention to Me"api0el Cameras 5egapi+el ca,eras are the wild card here. I# and when the total cost o# ownership =ca,era? ;andwidth? storage> o# ,egapi+el ca,eras gets close to analog ca,eras? the #inancial incentive to switch to I" could ;eco,e ver. strong. !ight now? it is hard to tell when and how that will ;e happening. However? i# .ou want to ;ene#it #ro, this transition? #ocus .our energies on understanding and anticipating this e,ergence. Security Mana"er Recommen$ations @or the 14 or 24K o# .ou that are alread. all I"? continue course. @or the rest o# .ou? .our decisions should ;e driven ;. two #actors: 1. Si2e o# the #acilities .ou ,anage: I# the. are s,all li8e Auic8 serve restaurants or ;outiAue retailers? ta8e .our ti,e with I"? no rush. I# the #acilities are large? .ou want to ,ove aggressivel. to I". 2. The state o# .our DV!: Chec8 the advances .our DV! supplier is ,a8ing. I# the. are ,a8ing advances li8e going h.;rid? supporting anal.tics? providing central ,anage,ent? etc.? .ou will li8el. ;e in good shape #or .ears to co,e. I# the. are not supporting this? .ou ,a. ;e ,issing out on this generationRs wave o# operational savings and loss reduction. In this case? start investigating ,igration to a new I" ;ased s.ste,.
Chapter 1+:
-op & Pro9le(s o IP Ca(eras
The overwhel,ing ,aEorit. o# people ;u.ing ca,eras toda. do not choose I" ca,eras. 3hile ,ost people see the ,ove to ;e inevita;le? serious de;ate e+ists on how long it will ta8e to get there. !ecentl. a nu,;er o# anal.sts have even slid ;ac8 proEections #or when I" ca,era sales will overta8e analog. To #igure out wh. .ou should select I" ca,eras and to #igure out when the ,ass o# users will ;u. I" ca,eras? we need to honestl. and clearl. assess what is holding ;ac8 I" ca,eras. The co,,on e+planation is a lac8 o# training and a lac8 o# support ;. traditional integrations and ,anu#acturers. I thin8 these pro;le,s are secondar. and a conseAuence o# ,ore #unda,ental pro;le,s. Here the. are: Primary Pro#lems

I" Ca,eras are too '+pensive Co,pared to -nalog Ca,eras Storage #or 5egapi+el ca,eras is too e+pensive S,art Ca,eras are still in their in#anc.
Secon$ary Pro#lems

DV!s o##er li,ited support 0ac8 o# Integrator Training
IP Cameras oo E0pensi'e It is co,,on 8nowledge that I" ca,eras cost ,ore than analog ca,eras. Co,pare a "elco CC%141H62 to a -+is 211. 3ith the sa,e lenses? the -+is is dou;le the price o# the "elco =L/44 to L%44>. However? this is not the ,ain point I want to ,a8e. To identi#. the real issues? we need to understand the #laws in -+is well pu;lici2ed TC& !eport on I" vs -nalog. The report clai,s that s.ste,s o# over (4 I" ca,eras are cheaper than analog. There is a lot in the report and it is certainl. worth reading. The #unda,ental pro;le, o# the report =and I" ca,era e+pense> is that over $4K o# custo,ers alread. have analog s.ste,s J DV!s? coa+ial ca;le? analog ca,eras? etc. The report assu,es a green#ield installation with no coa+ in place and no analog ca,eras. This is not representative #or the overwhel,ing ,aEorit. o# custo,ers. &nce .ou #actor in custo,ers that have analog in place and are loo8ing to upgrade? the costs loo8 #ar worse
#or I" ca,era s.ste,s. I would encourage -+is to conduct a #ollow6up report using this ,ore realistic assu,ption. 5oreover? ;eing *K cheaper than an e+isting technolog. =which is essentiall. what the -+is report clai,s> is rarel. su##icient to ,otivate custo,ers to switch. Indeed? in the applications where I" has ;een ,ost heavil. adopted? it is the eli,ination o# analog video #i;er networ8s that has provided the !&I necessar. to ,igrate to I" ca,eras. To ,a8e people switch? I" needs to ;e either signi#icantl. cheaper or o##er signi#icant econo,ic ;ene#its that analog can not o##er. -s I reviewed recentl.? I a, encouraged ;. the a;ilit. o# -CTi to provide solid? low cost I" ca,eras that can narrow the cost gap. &n the other #ront? to increase econo,ic ;ene#its? ,egapi+el ca,eras and video anal.tics hold the ,ost pro,ise. 0et<s now e+a,ine the,. Me"api0el Stora"e oo E0pensi'e Co,pared to analog ca,eras? ,egapi+el ca,eras can increase cases solved and reduce ca,era counts. The ;iggest econo,ic pro;le, #or ,egapi+el ca,eras is the cost o# storage. -l,ost all ,egapi+el ca,eras in production toda. use 5I"'G encoding which is 2+ to (+ less e##icient than the codecs used #or analog ca,eras. Co,;ine this with the ,assive increase in resolution o# ,egapi+el ca,eras and the cost o# storage per ca,era can ;e L*44 6 L1*44 per ca,era up #ro, L*4 6 L144 #or analog ca,eras. This is a signi#icant ta+ that ,an. custo,ers are Eusti#ia;l. concerned. Fou o#ten hear #ro, vendors that this is not a pro;le,. I ;elieve that #or certain high end custo,ers this is true =re,e,;ering less than 2K o# all ca,eras are ,egapi+el>. However? this will not ;e the case #or ,ainstrea, custo,ers. The e+tra cost #or storage will ,a8e it ver. hard to Eusti#. ,ainstrea, deplo.,ents o# ,egapi+el ca,eras. Fes? H.2/( is co,ing ;ut the Auestions on how well it will wor8 #or ,egapi+el ca,eras are signi#icant and unanswered. Smart Camera Infancy -nother wa. I" ca,eras can add value is ;. e,;edding video anal.tics into the ca,era. The challenge here is that s,art ca,eras are not co,,onl. availa;le =even #ro, -+is> and that ;ig Auestions re,ain a;out how well s,art ca,eras will wor8. 5a8ing this even ,ore challenging is that DV! ,anu#acturers are putting anal.tics directl. inside their units. This would e+tend the li#e o# analog ca,eras ,a8ing the case #or s,art ca,eras harder. DVRs Do Not Support
5an. DV!s o##er li,ited or no support #or I" ca,eras. This certainl. rein#orces the pro;le, #or I" ca,eras. 3hile ,an. I" vendors turn this into a ,oralit. pla.? the lac8 o# support is a re#lection o# a wea8 ;usiness case. I o#ten hear clai,s a;out the DV! co,panies are stupid or greed.. -t the heart o# it? I ;elieve the real ;arriers are #unda,entall. issues o# econo,ics J the pro;le,s I listed a;ove. &nce those pro;le,s are resolved? good DV! co,panies will ;e ;e ,otivated to support and ;ad ones that re#use will ;e Auic8l. crushed. (ac* of Inte"rator rainin" I see the sa,e issue #or securit. integrators as #or ,anu#acturers. Securit. integrators correctl. see that I" is not read. #or ,ost o# their custo,ers. -s such the ,otivation is wea8. -t the sa,e ti,e? securit. integrators are still doing well and growing. -gain? once I" ca,eras solve their pro;le,s? securit. integrators will ;e #orced to support or will ;e displaced. Conclusion one o# the a;ove ,eans that an. speci#ic custo,er should not use I" ca,eras. :se o# I" ca,eras depends on speci#ic application and logistic uses. However? I a, contending that until the % pri,ar. pro;le,s are solved I do not ;elieve I" ca,eras will ;e selected ;. a ,aEorit. o# custo,ers. The solutions o# these pro;le,s are:
Standard De#inition I" Ca,era onl. costs L144 ,ore than eAuivalent -nalog Ca,eras 5egapi+el ca,eras support H.2/( and H.2/( has no serious side e##ects on client or servers S,art Ca,eras are widel. availa;le and the anal.tics wor8 relia;l.
&nce these solutions are delivered? the ;usiness case will ;eco,e strong across the ;oard. Integrators and DV! ,anu#acturers will then ;e #orced to support I" ca,eras or ;e ousted ;. rivals that o##er the clearl. #inanciall. pre#era;le I" solution. Postscript @or an e+tensive discussion and de;ate on this topic? read the original report at http://ipvideo,ar8et.in#o/report/topT*TipTca,eraTpro;le,s.
Chapter 1.:
Value o H*9rid DVRs/NVRs
-l,ost all securit. ,anagers have DV!s. - ,inorit. have alread. ,oved to V!s and so,e still use VC!s ;ut 84K o# securit. ,anagers have DV!s toda.. -s such? what to do with .our DV!s and where to go ne+t is a ver. critical Auestion. H.;rid s.ste,s will ;e a 8e. part o# .our solution.
H.;rid V!/DV!s are appliances =purposed ;uilt co,puters> that can si,ultaneousl. support I" ca,eras and directl. connected analog ca,eras. This provides si,plicit. and #le+i;ilit.. Custo,ers can start with their e+isting analog ca,eras and slowl. ,igrate to I". Speci#icall.? unli8e a <pure< V!? a h.;rid V!/DV! eli,inates the need #or a separate video encoder when connecting to analog ca,eras. H.;rid V!/DV!s are now ;eing o##ered ;. al,ost all o# the traditional DV! co,panies. However? ,an. have Auestioned whether this ,eets a custo,er need or is done si,pl. ;ecause it is eas. #or the traditional DV! co,panies to do. evertheless? the h.;rid V!/DV! is Auite legiti,ate and pla.s a critical role in ver. co,,on scenarios in video surveillance:
84KM o# ca,eras toda. are analog and ,ost o# those ca,eras have ,an. .ears o# service le#t in the,. In ,an. applications =perhaps %4K or ,ore o# all s.ste,s>? ;andwidth constraints #orce custo,ers to deplo. recorders at the re,ote site near the on6site ca,eras.
In these scenarios? h.;rid V!/DV! s.ste,s will ;e ver. attractive. -nd since this scenario is ver. co,,on? it will ;e a ,aEor #actor #or ,an. securit. ,anagers and the industr. as a whole. To see wh. this will ;e a ,aEor #actor? let<s e+a,ine general V! ;ene#its and wh. the. are reduced in these scenarios. - ,ain ;ene#it o# a pure V! is consolidation o# video ,anage,ent and storage #unctionalities. !ather than ,anaging video in chun8s o# 1/ or %2 across potentiall. do2ens o# appliances? centrali2ed servers and storage clusters can ;e used. These servers and storage clusters can reduce eAuip,ent cost? power consu,ption and service costs.
Indeed? ,ain o# the earl. adopters o# pure V!s and I" video s.ste,s did so ;ecause o# this advantage. The ;iggest challenge in consolidation is ;andwidth availa;ilit.. Consolidating reAuires video #eeds #ro, various parts o# a #acilit./#acilities ;e trans,itted to a central location=s>. To do this? reAuires su##icient ;andwidth. Inside the local area networ8 =usuall. inside a ;uilding>? ;andwidth availa;ilit. is plent. and #airl. ine+pensive. However? in the wide area networ8 =usuall. ;etween ;uildings or ca,pus>? ;andwidth is scarce and Auite e+pensive. To centrali2e video ,anage,ent and storage across the 3could easil. cost hundreds or thousands o# dollars per ,onth? negating the ;ene#its o# consolidation. In ,an. distri;uted #acilities with ( to %2 ca,eras? organi2ations will have to ,anage and store their local #eeds in their local pre,ises. This is? o# course? not new as it is the co,,on practice with DV!s. However? it does a##ect the V! ;usiness case and create incentive to choose h.;rid V!/DV! s.ste,s.
Economic Comparison of +y#ri$ DVR>NVR to pure NVR 3hen .ou have less than %2 ca,eras and .ou need to store and ,anage those ca,eras locall.? the econo,ics o# h.;rid V!/DV!s are #ar ;etter than pure V!s. - ,id6tier 1/ to %2 channel h.;rid V!/DV! costs a;out L/?444 to L8?444 =using online Google pricing #or all esti,ates>. The h.;rid V!/DV! does encoding? storage? ,anage,ent and serving o# the video? all in one? with ,ini,al on6site setup and con#iguration. ). contrast? a pure V! solution can cost 24K J *4K ,ore than a h.;rid s.ste, and is ,ore co,ple+ to setup and ,aintain. The additional costs co,e #ro, having to =1> purchase standalone encoders to convert the analog ca,eras to I" =L244 to L%44 per ca,era>? =2> purchase so#tware licenses #or the V!=L144 to L1*4 per ca,era> and =%> purchase a "C/server with storage =L1* to L12* per ca,era>. -dditionall.? the server needs to ;e set up? so#tware loaded? &S tuned? encoders con#igured and connections esta;lished ;etween encoders and V!. It also ta8es ,ore space? ,ore I" addresses and ;ecause there are now ,ultiple s.ste,s? increases the ris8 o# integration or #uture service issues.
The V! approach is ,uch ,ore co,ple+ and ti,e consu,ing than the co,parative h.;rid V!/DV! which is relativel. plug and pla.. In a large scale environ,ent where 144s o# ca,eras were ;eing consolidated? the cost savings o#ten Eusti#. the additional co,ple+it. and setup ti,e. However? in a s,all setup? the costs are Auite signi#icant.
+y#ri$ DVR>NVRs Pro'i$e a Smoot% ransition @or an. given custo,er? the ,ost attractive h.;rid DV!/ V! will ;e the unit #ro, their e+isting DV! supplier. 'ven i# the custo,er does not especiall. li8e their DV! vendor? all o# their sta## is trained on using that DV!<s client so#tware. 5oreover? o#ten? all o# the DV!s are #ro, one vendor? so the sta## never has to worr. a;out which so#tware client to use. The sa,e client so#tware #or the DV! can usuall. ;e used #or the h.;rid s.ste,s. This ,a8es the switch sea,less and transparent to the users. Custo,er are willing to switch ;ut when it<s close? the co,#ort o# the sta## is a ,aEor #actor in stic8ing with e+isting processes and products.
-%at6s t%e Downsi$e of +y#ri$ DVR>NVRs The ;iggest downside o# H.;rid DV!/ V!s is that ,an. are not trul. h.;rid. - genuine h.;rid would ;e eAuall. #le+i;le with I" and analog. 5i+ing and ,atching ,an. co,;inations o# analog and I" would ;e standard. Supporting a variet. o# I" and ,egapi+el ca,eras would also ;e standard. '+acA is a good e+a,ple o# a true h.;id. The pro;le, is a lot o# so called <h.;rid< s.ste,s o##er onl. to8en support #or ,i+ing and ,atching and #or di##erent I" ca,eras. &ne co,,on techniAue is to o##er onl. a #ew additional I" ca,eras? constrained to 1 or 2 I" suppliers? in addition to the 1/ analog inputs. G'<s S.,dec is an e+a,ple o# a B#a8eB h.;rid. H.;rid s.ste,s are supposed to give .ou #le+i;ilit. to grow into I". This approach is ,ore o# a tric8 than a ;ene#it.
ew &ptions #or DV!/ V! Storage
Chapter 10:
New "ptions or DVR/NVR Storage
@or .ears? storage #or video surveillance has ;een done on ;oard .our DV!. Fou speci#ied the si2e o# the hard drive .ou need? the ,anu#acturer ,ade sure the right hard drives were installed and .our unit was shipped to .ou. Toda.? a ,aEor new option is e,erging that replaces storage in .our DV!s/Servers and places the, in central clusters o# storage.
In the last .ear? ;u22 and vendor ,ar8eting has grown Auic8l. around clustered storage solutions that could replace the traditional internal storage that has ;een the standard #or ,an. .ears in video surveillance. Despite earl. wins ;eing concentrated in a #ew niche ,ar8ets =e.g.? casinos? ,unicipalities>? the #unda,entals o# these o##ering indicate the. will have a ,aEor i,pact across ,ost o# the video surveillance industr. in the ne+t % .ears. This review e+a,ines the ;ac8ground? advantages and constraints o# storage clusters as a replace,ent #or traditional DV!/ V! storage. Storage clusters are appliances that are separate #ro, .our DV!/ V! and co,,unicate with the, across .our I" networ8. Storage clusters are ,odular and ,ore storage can ;e added over ti,e? starting #ro, as low as a #ew T)s to ,ore than 1444 T)s. The ,ost well 8nown specialists o##ering these solutions are Intransa and "ivot% . Recommen$ation5 :se storage clusters when a site has ,ore than N (8 analog ca,eras and/or ,ore than N / ,egapi+el ca,eras. Here is a su,,ar. o# the 8e. advantages and constraints:
I"Video5ar8et.In#o
8$
A$'anta"es 1. "rice di##erential ;etween internal storage and storage clusters have dropped dra,aticall. 2. Storage Clusters can reduce storage needs ;. N %4K over internal storage %. Storage Clusters are actuall. cheaper #or large ca,era counts and storage durations (. Storage Clusters are cheaper and ;etter #or ,egapi+el ca,eras *. Storage Clusters o##er !-ID <standard< Constraints 1. Storage Clusters are not cost6e##ective #or s,aller ca,era counts 2. Storage Clusters cannot centrali2e storage across ,ost distri;uted #acilities
A$'anta"e 45 Price Differential 3here historicall. the price di##erential per unit o# storage was huge? toda.? the di##erential is s,all or? in ,an. cases? not at all. This is critical in spurring ;roader adoption. -l,ost all o;servers recogni2ed that storage clusters were superior to on6;oard storage ;ut the historical pricing #or storage clusters was %44K to /44K ,ore #or the clusters. -s such? it was incredi;l. di##icult to Eusti#. the signi#icant increase in e+pense and ver. #ew video surveillance s.ste,s used this solution. In the past #ew .ears? the increasing ,aturit. o# these solutions and the utili2ation o# standards ;ased I" networ8s has shrun8 the price di##erential. The price o# the supporting in#rastructure to ;uild storage clusters has dropped? ena;ling the price o# storage clusters to #all #aster than the price o# internal DV!/ V! storage. "er T)? the price o# storage clusters is ver. co,petitive with internal DV!/ V!
I"Video5ar8et.In#o
$4
storage. Storage clusters cost a;out L2?444 per T). @or ,ost ,ainstrea, DV!s? the 5S!" to add 1 T) o# storage is L2*44 to L%844. Storage clusters actuall. can ;e cheaper than internal storage. I was #airl. shoc8ed a;out this di##erence ;ut I con#ir,ed with ,ultiple price lists #ro, ,ultiple dealers. ote: The ,ini,u, si2e storage cluster availa;le toda. is (T) which is a ;ig #actor in s,all ca,era count deplo.,ent. This a##ects total cost #or this scenario and will ;e e+a,ined in the constraints section.
A$'anta"e 75 Re$uce Stora"e Nee$s 3hen .ou deplo. ,ultiple DV!s? even in the sa,e ;uilding and with the sa,e con#igurations? .ou o#ten o;tain di##erent storage durations. 0et<s sa. .ou target $4 da.s o# storage? so,e will get 1*? others 14*? a #ew ** and one or two 12* da.s. This is ;ecause storage utili2ation is a #actor o# a,ount o# ,otion or tra##ic in a ca,era<s view. ="THs are #a,ous #or chewing up storage ;ecause o# this>. )ecause storage duration #alls in a range? .ou generall. need to deplo. ,ore storage than the average s.ste, needs or reducing recording settings on s.ste,s that do not record as long as .ou need. In the #or,er? the direct i,pact is higher installation costs. In the later? the direct i,pact is higher service costs and the indirect i,pact can ;e issues with evidence usa;ilit.. 3ith storage clusters? DV!s/ V!s record to a central pool o# storage. 0et<s sa. .ou have 14 DV!s and the average DV! consu,es 844 G)s o# storage to record $4 da.s. However? ;ecause so,e o# the DV!s will need ,ore storage to reach $4 da.s? .ou use 1444G) storage in each DV! instead? resulting in a 2*K pre,iu,. In a storage cluster? ;ecause storage is pooled? units that need ,ore storage and ;alanced o## with units that need less. There#ore? .ou would si,pl. use 8T) o# storage and eli,inate the 2*K padding and pre,iu,.
I"Video5ar8et.In#o
$1
'ven in a ,odest scenario li8e this .ou can save a #ew thousand dollars si,pl. ;. this pooling e##ect.
A$'anta"e 85 C%eaper for (ar"e Camera Counts > E0ten$e$ Stora"e &nce .ou get into large ca,era counts or e+tended storage? it pushes the li,its o# what internal storage can provide. -s such? the econo,ics o# storage clusters are pre#era;le. :sing interal storage will reAuire using the largest hard drives possi;le =so .ou can #it in chassis>. 0arger hard drives are ,uch ,ore e+pensive =per unit o# storage> than s,aller hard drives. 3ith a storage cluster? .ou could use the ,ost cost6e##ective hard drives si2es and reduce costs. :sing directed attached storage reAuires an e+ternal appliance. Such an appliance will cost a #ew thousand? even without the drives. -t that point? and #or roughl. the sa,e price? .ou ,ight as well use a storage cluster that provides #ar superior scala;ilit..
A$'anta"e 95 Better for Me"api0el Cameras The storage de,ands o# ,egapi+el ca,eras are severe. Fou can easil. use ,ultiple T)s per ,egapi+el ca,era ;ecause o# the increased resolution and the need #or 5I"'G co,pression. 5ost DV!/ V! appliances are not designed to handle this de,and #or storage co,ing #ro, ,ultiple ,egapi+el ca,eras. Fou can use direct attached storage ;ut again? #or roughl. the sa,e price? the added ;ene#its o# a storage cluster generall.
I"Video5ar8et.In#o
$2
,a8e it the appropriate choice.
A$'anta"e :5 Offers RAID 6stan$ar$6 3hile DV!s have o##ered !-ID #or .ears? the additional cost has ;een Auite high. -s such? ,ost users elect not to use !-ID. - chec8 o# leading appliances indicates going #ro, 1T) non6!-ID to !-ID increases cost ;. L1*44 to L%444 5S!". So,e s.ste,s li8e %V! are now o##ering !-ID standard ;ut it<s certainl. ,ore o# an e+ception than the nor,. -ll in all? though? these pre,iu,s are hard to Eusti#.. ). contrast? storage clusters o##er !-ID <standard<. The incre,ental cost o# using !-ID is ver. low. "lus? these s.ste,s are all designed to provide ,an. #lavors o# !-ID right out o# the ;o+. !-ID o##ers two pri,ar. ;ene#its #or video surveillance s.ste,s: =1> save the ;o+ #ro, d.ing and =2> save video #ro, ;eing lost. Saving the ;o+ #ro, d.ing is the ,ore i,portant o# the two. Video surveillance s.ste,s that die reAuire e,ergenc. service calls and increase the ris8 that a securit. incident occurs where no recording nor live ,onitoring is availa;le. Saving video #ro, ;eing lost is usuall. a nice to have #or video surveillance. It is generall. not critical ;ecause the pro;a;ilit. o# a incident ;eing lost is usuall. low. This is ;ased on ,. e+perience in govern,ent? ,ilitar. and @ortune 2444 co,panies. Video storage is generall. ,ore li8e insurance than it is li8e .our children. I# .ou lose .our insurance #or a #ew da.s? it<s a ris8 ;ut .ou are usuall. #ine. Video surveillance storage is prett. ,uch the sa,e.
I"Video5ar8et.In#o
$%
Certainl.? a lot o# reasons e+ist #or ,oving to storage clusters. evertheless? the value #or s,aller? distri;uted sites =a ,aEor seg,ent o# video surveillance 6 ;an8s? s,aller retailers? QS!s> is not as strong. Constraint 345 Not Effecti'e for Smaller Camera Counts Storage clusters have a startup cost that is nota;l. higher than the internal storage DV!s/ V!s use. Iust li8e an. "C? internal storage is availa;le ;. de#ault and the onl. incre,ental cost is usuall. adding in the drives the,selves. 3ith a storage cluster? .ou usuall. have a separate appliance with electronics and co,puting in#rastructure. -s such? ;e#ore .ou deplo. an. drives with a storage cluster? .ou #irst have to pa. #or this additional appliance. 3ith a storage cluster? the ,ini,u, availa;le si2e see,s to ;e (T) at a;out L8?444 5S!". I# .ou use signi#icantl. less than (T)? the cost #or a storage cluster will ;e signi#icantl. higher than si,pl. adding in drives to .our DV!/ V!. 5oreover? the ;ene#its o# pooling decrease ;ecause .ou are li8el. using less DV!/ V! units and do not need to worr. a;out padding to achieve storage durations. This is not trivial ;ecause large corporations have ,illions o# #acilities around the world that #it these characteristics. :ntil and i# storage clusters can ;eco,e ,ore co,petitive at lower storage entr. levels? the value #or custo,ers will ;e Auite Auestiona;le.
Constraint 375 Not Capa#le of Centrali1ation Across Distri#ute$ &acilities 'ven though storage clusters have signi#icant econo,ic ;ene#it #or large a,ount o# storage use? this is generall. not #easi;le #or aggregating storage #ro, #acilities across the countr.. To the credit o# the vendors in this space? the. have not ,ade
I"Video5ar8et.In#o
$(
this clai,. However? .ou do hear this #ro, ti,e to ti,e ;. so,e in the industr.. Ta8e a #ast #ood restaurant with 1444 locations. In each location the. pro;a;l. have a;out 2*4 G) o# storage =,ore i# the. are rolling out new s.ste,s>. In total? that<s 2*4 T) o# storage =at least>? which is Auite signi#icant. H.potheticall.? the. could save a #ew hundred thousand dollars i# the. could eli,inate hard drives in the restaurants and Eust have one central storage cluster. The pro;le, is that .ou need signi#icant a,ounts o# ;andwidth to acco,plish this that si,pl. is not availa;le to ,ost sites. @or 8 or 1/ ca,eras? .ou ,ight need * 6 24 5;/s in upstrea, ;andwidth. This is a huge a,ount #or ,ost stores where DS0/ca;le ,ode, is the nor,. In other words? the cost o# ;andwidth is #ar higher than the cost o# storage. -s such? this is not ver. realistic.
Conclusion Driven ;. price co,petitiveness and a nu,;er o# signi#icant advantage? storage clusters will Auic8l. ;eco,e a ,aEor #orce in sites with ,odest to large nu,;ers o# ca,eras. evertheless? the si2a;le seg,ent o# the ,ar8et with s,all ca,era counts per site will not see signi#icant advantages in this. -ll integrators and securit. ,anagers should care#ull. trac8 and learn ,ore a;out securit. clusters so the. can ta8e utili2e their signi#icant advantages.
I"Video5ar8et.In#o
$*
The Value o# Centrali2ed Video -nal.tics
Chapter 12:
-he Value o Centrali/ed Video )nal*tics
"er#or,ing anal.tics at the DV!/ V! is a ver. popular choice in video surveillance. !eading the trade ,aga2ines and viewing ,ar8eting ,aterials? .ou pro;a;l. would not thin8 so ;ecause $4K o# the attention goes to anal.tics at the edge. To ,e? this is a classic case o# what vendors want the world to ;e rather than what is ;est #or custo,ers. @or ,an. custo,ers? centrali2ed anal.tics are a #ar ;etter choice. Here are the top reasons:

!euses .our e+isting ca,eras. 0ess e+pensive than anal.tics at the edge. Si,pli#ies ,aintenance and upgrade o# anal.tics.
Reuse E0istin" Cameras Custo,ers? Eusti#ia;l.? hate to throw awa. usa;le assets. The pro;le, is to use s,art ca,eras .ou will generall. have to throw awa. e+isting assets or purchase new assets that are redundant. &ver 84K o# ca,eras deplo.ed are analog 6 none o# the, support anal.tics. Fou will either have to replace the, or add an encoder with ;uilt6in anal.tics. @urther,ore? ,ost o# the I" ca,eras do not have the capa;ilit. to ;e upgraded to video anal.tics. -s such? even i# .ou deplo.ed new I" ca,eras / ,onths ago? those too would have to ;e replaced =perhaps .ou can relocate or resell those ca,eras>. Fou ,a. ;e in a situation where .ou are adding new ca,eras or are read. to throw awa. e+isting ca,eras. In that case? great. This is not a pro;le,. )ut ,ost people are not in that situation. (ess E0pensi'e
I"Video5ar8et.In#o
$/
The Value o# Centrali2ed Video -nal.tics
Consolidating video anal.tics at the DV!/ V! reduces cost. S,art ca,eras are usuall. a #ew hundred dollars ,ore than traditional I" ca,eras. S,art DV!s/ V!s =s.ste,s that consolidate anal.tics and video recording in one> are usuall. no ,ore than a thousand dollars ,ore than a DV!. Here<s what happens. I# .ou ;u. a s,art DV!/ V! .ou pa. a;out a thousand dollars ,ore. I# .ou ;u. s,art ca,eras and need to replace .our e+isting ca,eras? .ou can easil. pa. L*?444 to L1*?444 ,ore. 'ven i# .ou do not have to throw awa. e+isting ca,eras? it can cost .ou a #ew thousand dollars #or the pre,iu, o# s,art ca,eras. @airl. s,all co,panies li8e Clic8it and i%DV! are ;eing selected ;. ,an. custo,ers #or Eust this reason. The. have ;uilt DV!s with anal.tics inside that onl. cost ,odestl. ,ore than a traditional DV!. The econo,ics are such that this is o;viousl. ver. attractive and despite the #act that these co,panies do not have ;ig ,ar8eting dollars? there are getting so,e i,pressive wins. Simplifies Maintenance an$ =p"ra$e ot a lot o# people are tal8ing a;out this .et ;ut ,aintenance and upgrade are going to ;e ver. pro;le,atic with s,art ca,eras. S,art ca,eras are shin. toda. ;ut in %6* .ears? .our s,art ca,era is going to loo8 li8e the #irst generation i"od 6 ;ig? ;ul8.? stupid and unco,petitive with the 2411 state o# the art. It<s not practical to replace the chips inside each ca,era where it is ver. practical to upgrade a centrali2ed server than handles anal.sis #or nu,erous ca,eras. Conclusion 3hile I a, not opposed to anal.tics at the edge? the radical i,;alance o# coverage and treat,ent o# this issue is a disservice to securit. ,anagers. Hope#ull.? this article helped to shed so,e light on the o#t6overloo8ed ;ene#its o# centrali2ed anal.tics.
I"Video5ar8et.In#o
$1
'+a,ining <&pen< S.ste,s
Chapter 23:
#'a(ining :"pen: S*ste(s
3hile ;eing <open< is the trend? <openness< is vague? clai,ed ;. all and underesti,ated in its di##icult. to achieve. I# .ou are ;u.ing or speci#.ing video ,anage,ent s.ste,s? .ou need to care#ull. consider this. ot too long ago? I was sitting with one o# the ,ost 8nown and respected e+perts in CCTV. He e+pressed his #rustration and dis,a. that a vendor who told hi, the. were open were actuall. not. This was having a serious i,pact on s.ste,s he was designing. ow? i# he could get caught ;. this? this could happen to an. o# us. Here are the top % pro;le,s I see:

B&pennessB is vague 6 what does it actuall. ,ean9 'ver.one clai,s to ;e open 6 even i# the. are not reall. )eing open is hard ;ut it<s routinel. assu,ed as eas.
)ecause o# this? .ou ,a. never 8now the truth and ;e stuc8 with a s.ste, that is loc8ing .ou in. Openness is Va"ue -t a ;asic level? ;eing open ,eans that a s.ste, can wor8 with other s.ste,s #ro, di##erent ,anu#acturers. )ut how ,an. other s.ste,s should a s.ste, wor8 with to ;e called open9 -nd how ,an. other ,anu#acturers do .ou need to wor8 with to ;e called open9 !espected industr. leaders o#ten de#ine openness as a vendor wor8ing with one or two other ,anu#acturers in a single categor.. Certainl. this is so,ewhat open ;ut is it open enough9 @or ,ost users? it is not and poses a ;ig ris8 that when the da. co,es #or .ou to integrate with a di##erent s.ste, or product that it Eust will not wor8. I"Video5ar8et.In#o $8
E'eryone Claims to #e Open To ,e? this is the ,ost dangerous ele,ent in the <openness< discussion. "oliticians have learned that racis, is no longer accepta;le. So is the result that no politician is racist an.,ore9 &# course not. The result is that politicians 8now to avoid racist language and ,a8e clai,s to racial eAualit.. This is analogous situation with video surveillance s.ste,s. !egardless o# how closed a s.ste, is? all sales and ,ar8eting people 8now that .ou ,ust clai, to ;e open regardless o# how open .ou reall. are. To pu;licl. state to a client that .ou are not open is ver. ris8. so to solve that pro;le, vendors si,pl. clai, that the. are open. -nd ;ecause the co,,onl. accepted de#inition o# openness is so vague? it<s eas. to do it without reservation. Openness is +ar$ It see,s as i# vendors si,pl. will openness into e+istenceP -s i# the act o# sa.ing .our open ,a8es .ou open. It<s ;ac8ed up ;. the a;surd clai, that B3e have an -"I.B Though .ou need an -"I? si,pl. having an -"I is Eust the ;eginning. It<s li8e sa.ing .our are a Che# ;ecause .ou can ;ar;ecue ha,;urgers. The realit. is that trul. ;eing open ta8es a huge co,,it,ent #ro, the vendor. It ,eans opti,i2ing .our -"I to ,a8e it easier #or other parties to use. It ,eans doing custo, integrations to support other people who use legac. technologies or are not as open. -nd perhaps ,ost o# all it ,eans a huge develop,ent e##ort to actuall. support the hundreds o# devices out there. &ne o# ,. #avorite Auestions to as8 is? B3hat products do .ou actuall. support toda.9B This s,o8es out a lot o# spin and h.pe o# <open s.ste,s.< 5ost vendors ta8e the approach that i# it<s theoreticall. possi;le #or the, to integrate with another product that the. can clai, to a custo,er that the. support the product. )eware o# this. "ush #or the details and s,o8e out the truth. Conclusion
I"Video5ar8et.In#o
$$
-s a #irst step? we all need to ;e care#ul a;out properl. assessing openness. I also thin8 we ,a. need to start getting ;etter de#initions and assess,ents o# how open s.ste,s are.
I"Video5ar8et.In#o
144
The Danger o# )u.ing "ac8ages
Chapter 21:
-he Danger o %u*ing Pac6ages
- dangerous and ,ass ,ove,ent is underwa. #or video surveillance co,panies to sell .ou pac8ages. "ac8aging together ca,eras? encoders and I" video ,anage,ent s.ste,s? vendors hope to entice .ou with an integrated? opti,i2ed end to end solution. &ne o# the great ironies is that while ever.one is pa.ing lip service to open palt#or,s? the industr. is clearing ,oving to ,ore tightl. ;undled pac8ages. I thin8 this is ver. ris8. and .ou should care#ull. consider the dangers o# ;u.ing Bsolutions.B Bpac8agesB. &riginall. I called this solutions. I ;elieve this is a poor choice o# words. I have now changed to pac8ages to ;etter connote the pheno,enon.
3ho<s selling pac8ages9 Verint? 5arch? -,erican D.na,ics? "elco? Cisco? DvTel? )osch? IndigoVision? -vigilon. Fou can even see the ;eginning o# this with -+is with their e+pansion o# Ca, Station. Toda.? "anasonic announced it to was ,oving to selling BsolutionsB? i.e pac8ages. It<s al,ost easier to as8 who is not selling pac8ages =5ilestone ;eing the ,ost o;vious large pla.er>. -nd what<s 8e. is that? * .ears ago? a lot o# these co,panies speciali2ed in Eust ,anage,ent s.ste,s or ca,eras. The trend is e+panding. Vendors love the thought o# selling pac8ages ;ecause it has the potential to increase revenue =;. cross selling> and to increase ,argins =;. ;undling>. The. can also tell the,selves that the. have ,oved up ,ar8et and are delivering greater value? etc? etc. I do not dou;t that so,e vendors can ;ut when .ou have ,ore than a do2en vendors all selling #unda,entall. the sa,e pac8age? .ou have a ver. ris8. situation #or ever.one involved. I"Video5ar8et.In#o 141
The Danger o# )u.ing "ac8ages
Dan"er 45 Pac*a"es Are oo )eneral Video surveillance ;u.ers have a wide variet. o# needs. However? ,ost pac8ages are hori2ontall. positioned =that is? the. are not opti,i2ed #or an. speci#ic use case>. "ac8ages can restrict #le+i;ilit. and adapta;ilit. to di##erent use cases. )e care#ul that the pac8age properl. addresses .our need. Dan"er 75 ;ou are screwe$ if you c%oose a Mar*et (a""in" Pac*a"e Since there are so ,an. vendors selling pac8ages? so,e o# the, are going to lose. Fou cannot e+pect to have a do2en co,panies all ;asicall. o##ering the sa,e thing to all succeed. I# .ou choose a pac8ages that loses? .ou are in trou;le. It will ;e ver. hard to e+pand the pac8age and .ou will li8el. ;e loc8ed in to its li,itations. Dan"er 85 ;ou are controlle$ if you c%oose a Mar*et (ea$in" Pac*a"e I# the pac8age wins? .ou ;eco,e at the ,erc. o# the vendor. This is wh. so ,uch ill will e+ists towards co,panies li8e G' Securit. and T.co. The. got .ou into their pac8age and the. 8now it. !eAuests #or supporting third part. products or new #eatures are slow or unli8el. to ;e approved even i# .ou are a giant custo,er. This ,ove to I" video solutions see,s to ris8 replicating the sa,e pro;le,s we have ;een struggling with #or the last decade. I a, not sa.ing .ou should not ;u. pac8ages. I thin8 so,e o# the, are particularl. strong =especiall. to the e+tent the. #ocus on a vertical>. However? .ou should clearl. understand these ,oves and #actor in the ris8s o# the,.
I"Video5ar8et.In#o
142
Introduction to Cit.63ide Surveillance
Chapter 22:
Introduction to Cit*51ide Surveillance
3hile the pu;licit. goes to ,ega6cities deplo.ing video surveillance? cit. wide video surveillance is ;eco,ing ver. cost6e##ective? valua;le and via;le #or <nor,al si2ed< cities throughout the world. The ,edia coverage #or the ew For8 and Chicago cit. deplo.,ents can ,a8e cit.6wide video surveillance see, daunting. It appears that .ou need:

Tens o# ,illions in #unding Ho,eland securit. grants De#ense contractors Deplo.ing co,ple+ new wireless networ8s Integrating ,ilitar. st.le co,,and and control s.ste,s
In realit.? though? ,ost an. cit. can ;ene#it #ro, ,ature? ine+pensive and #le+i;le cit. wide video surveillance. The Cit. o# 0ong,ont? Colorado provides a nice e+a,ple o# how to deplo. cit. wide video surveillance. &ver the last #ew .ears? 0ong,ont =population 11?444> has developed a #irst class cit. wide solution. 3or8ing with Volpe Industries and using video surveillance solutions #ro, -+is Co,,unications and Dete+i S.ste,s? 0ong,ont now has a power#ul surveillance solution. This article e+plores so,e o# the 8e. principles used in this deplo.,ent to ;uild a success#ul cit. wide surveillance s.ste,. Principles for Success5 1. :se '+isting I" etwor8s that the 5unicipalit. 5anages 2. Deplo. V! servers at @acilities Throughout the @acilit. %. :se Direct3ireless 0in8s to Connect Ca,era to @acilities (. Share Video #ro, the Cit. with the "olice Depart,ent<s Co,,and Center I"Video5ar8et.In#o 14%
*. '+pand the S.ste, Step ;. Step? Fear ;. Fear
Principle 45 =se E0istin" IP Municipality Networ*s 5ost cities ,anage their own internal networ8s #or cit. #acilities. These networ8s are a si,ple and low cost wa. to ;uild a cit. wide surveillance s.ste,. These networ8s generall. connect cit. #acilities across the ,unicipalit. and o#ten provide high speed connectivit.? to ;oost. The integrator? Volpe? wor8ed closel. with the 0ong,ont<s CI& to put the I" video surveillance s.ste, on the ,unicipalities networ8. The incre,ental cost o# adding the s.ste, was practicall. nil as it leveraged this e+isting in#rastructure. Principle 75 Deploy NVR se'ers at Eac% &acility @or each #acilit. using I" Video Surveillance =le.g.? i;raries? schools? cit. hall? police stations? etc>? an V! server was deplo.ed to store and ,anage video #ro, the local ca,eras. This ta8es advantage o# video ;eing recorded ,uch ,ore o#ten than it is viewed. 3hile video ,ust ;e strea,ed #ro, the ca,eras #reAuentl. or constantl. to ena;le recording? onl. occasionall. does the video need to ;e viewed ;. cit. o##icials such as the police. "lacing an V! on each site signi#icantl. reduces the a,ount o# ;andwidth needed on the ,unicipalities networ8 that connects di##erent #acilities. This is i,portant in ,a8ing the addition o# video surveillance as low i,pact as possi;le on the ,unicipalities networ8s. 3hile centrali2ing video storage o##ers so,e savings in server and storage consolidation? li8e ,ost organi2ations? 0ong,ont ;ene#ited ,ore #ro, ,ini,i2ing the i,pact on the cit.<s networ8. Principle 85 Direct -ireless (in*s from &acilities - nu,;er o# outdoor locations needing surveillance were econo,icall. addressed through direct wireless lin8s originating #ro, cit. #acilities. @ro, the roo# o# cit. #acilities? wireless lin8s were esta;lished to various points o# interests within a I"Video5ar8et.In#o 14(
#ew ,iles #ro, the #acilit.. -s these connections were direct =point to point or point to ,ultipoint>? the. were rather ine+pensive and si,ple to esta;lish. Principle 95 S%are Vi$eo The hundreds o# ca,eras now deplo.ed across the cit. can now ;e leveraged ;. the police to help respond to e,ergencies or critical investigations. ). using the cit.<s I" networ8? the "olice can access an. o# the V! servers in the cit. to view live or recorded video. This has alread. helped handle real ti,e incidents as well as solve cases where cri,inals have ,oved across the cit.. Principle :5 E0pan$ Step #y Step 3hile cit. wide surveillance is o#ten thought o# as a ,assive proEect? 0ong,ont de,onstrates that the s.ste, can grow incre,entall.. Indeed? at 0ong,ont? new #acilities have ;een co,ing on line ever. .ear #or the last couple o# .ears. This allows cit. wide surveillance s.ste,s to grow in s,all chun8s reAuiring tens o# thousands o# dollars rather than ,illions o# dollars. ). leveraging the cit.<s 0- and deplo.ing surveillance at cit. #acilities and their surroundings? step ;. step? the cit. can grow the s.ste,. This can ;e Auite valua;le in ;udget allocations and in ;uilding pu;lic support #or the value o# the s.ste,. &uture Consi$erations Cit. wide video surveillance s.ste,s can and will continue to grow. 3hile 0ong,ont is a great e+a,ple o# leveraging .our e+isting resources and starting with si,ple? high value uses? cit. wide video surveillance has the potential to e+pand to address other concerns. )elow are an e+a,ple o# those #uture considerations:
Inter#acing with 0ocal )usiness V!s/DV!s: - lot o# interest e+ists in sharing video ;etween cities and local ;usiness to help respond to cri,es or disasters. This reAuires support or interaction with ,an. di##erent ,anu#acturers o# V!s/DV!s. Currentl.? 0ong,ont? li8e ,an. 14*
I"Video5ar8et.In#o
,unicipalities has standardi2ed on a single V! plat#or, #ro, Dete+i. This provides great si,plicit. as all ca,eras are ,anaged ;. this s.ste,. evertheless? it is co,,on that one ,anu#acturer<s V! cannot access or view ca,era<s #ro, another V! s.ste,. Co,,and and Control s.ste,s can ;e put in place to access ,ultiple V! s.ste,s ;ut this can signi#icantl. increase cost and co,ple+it..
:sing on6Cit. I" etwor8s: So,e #acilities and so,e points o# interest are not served ;. the Cit.<s I" etwor8. In those cases? the use o# DS0 or ca;le ,ode, #ro, the local teleco,,unications providers will ;e necessar.. This will reAuire so,e added co,ple+it. to provide securit.? etc. ;ut is increasingl. ;eco,ing an accepted and straight#orward part o# cit. wide surveillance s.ste,s. "rovide )roader Ca,era Coverage: In so,e cases? cities want to deplo. ca,eras in places where there are no wired networ8s and direct wireless connections are not #easi;le. 3ireless ,esh networ8s are an ideal solution to this pro;le,. The. are designed to handle o;stacles? provide greater redundanc. and cover ;roader areas. evertheless? the. also are ,ore co,ple+ to design and ,ore e+pensive to deplo..
Conclusion 3ith I" video? cit. wide video surveillance is ;eco,ing an a##orda;le and valua;le wa. to i,prove securit.. ). #ollowing the principles o# 0ong,ont<s deplo.,ent? ,ost cities can Auic8l. and #airl. si,pl. roll out cit. wide video surveillance s.ste,s that provide a strong #oundation #or continuous i,prove,ent and new ;ene#its.
I"Video5ar8et.In#o
14/
Is "u;lic CCTV '##ective9
Chapter 2!:
Is Pu9lic CC-V # ective7
3hile we continue to spend ,ore on pu;lic CCTV s.ste,s? the de;ate on CCTV e##ectiveness has reached a polari2ing and inconclusive stando##. &n the one side? .ou have a nu,;er o# studies and leading thin8ers who clearl. contend that CCTV s.ste,s are ine##ective. &n the other? .ou have nu,erous ,unicipalities who are wee8l. green6lighting new CCTV proEects. This report o##ers 8e. #indings #ro, the 24 top studies/articles in the #ield and o##ers practical reco,,endations on how to opti,i2e the use o# pu;lic CCTV s.ste,s. Bey &in$in"s Summary
The e+pectation that CCTV s.ste,s should ;e deplo.ed to reduce cri,e rather than solve cri,e has created huge pro;le,s. 3hile the studies show serious dou;t on CCTV<s a;ilit. to reduce cri,e generall.? a strong consensus e+ists in CCTV<s a;ilit. to reduce pre,editative/propert. cri,e CCTV is consistentl. treated as a singular? sta;le technolog.? o;scuring radical technological changes that have occurred in the last 14 .ears Di##erences in per ca,era costs are largel. ignored? preventing polic. ,a8ers #ro, #inding wa.s to reduce costs !outine co,parison o# police vs ca,eras is counterproductive
Practical Recommen$ations Summary

Stop clai,ing that CCTV can generall. reduce cri,e &pti,i2e #uture pu;lic CCTV proEects around cri,e solving rather than 141
I"Video5ar8et.In#o
cri,e reduction
&pti,i2e #uture pu;lic CCTV proEects around ,aterial and pre,editative cri,es Target technologies that support cri,e solving and ,aterial/pre,editative cri,es @ocus on ,ini,i2ing cost per ca,era
&in$in"5 Crime Re$uction 's Crime Sol'in" The overwhel,ing ,aEorit. o# studies #ocus on anal.2ing CCTV<s a;ilit. to reduce cri,e. The general approach is to ta8e current cri,e statistics #or a region and co,pare those statistics to the period a#ter installation o# CCTV. - nu,;er o# techniAues are used to adEust to control #or general changes in cri,e and to trac8 displace,ent or di##usion o# ;ene#its to other areas. evertheless? the #ocus o# all Auantitative anal.sis has ;een on reducing cri,e. This is the ,irror opposite o# the private sector. In the private sector? the overwhel,ing ,aEorit. o# CCTV s.ste,s are Eusti#ied ;. their use in solving cri,e. It is investigations where ,ost private ;usinesses #ind value and return in their CCTV s.ste,s. @or ;usinesses? onl. a ver. s,all percentage o# CCTV ca,eras are ever even watched. The s.ste,s pa. #or the,selves ;. periodicall. ;eing a;le to identi#. or prove a cri,inal activit.. This indicates a #ailure o# e+pectations #or pu;lic CCTV s.ste,s. In the private sector? when CCTV e##ectiveness is discussion? the assu,ption is usuall. that CCTV is used #or investigations. ). contrast? the #ocus on pu;lic CCTV e##ectiveness ;eing deter,ined on reducing cri,e sets a dangerous e+pectation that is di##icult to achieve and li8el. to create dissatis#action within the co,,unit.. The pro;le, see,s to ;e the #ault o# the original advocates o# these s.ste,s? rather than a de#icienc. o# the testers. The acade,ics and researchers per#or,ing I"Video5ar8et.In#o 148
these tests were reacting to the e+pectations that the proponents o# these s.ste,s ,ade originall.. In the reco,,endations section? I will e+a,ine how we can ,ove ;e.ond this unproductive and pro;le,atic situation. &in$in"5 Re$ucin" Crime )enerally 's Preme$itati'e>Property Crime The ,edia<s ,ain #ocus has ;een on whether or not CCTV reduces cri,e as a whole. This o#ten has turned the issue into an all or nothing de;ate. The testing has also #ocused on the general i,pact on cri,e reduction ;ut nota;le attention has ;een paid to di##erent t.pes o# cri,e. 3idespread consensus e+ists that CCTV is e##ective in reducing pre,editative/propert. cri,e. -ll the studies ac8nowledge this? even the -C0:<s which otherwise is e+tre,el. negative towards CCTV. The ,ost #reAuentl. cited e+a,ple is the a;ilit. to reduce the#ts in par8ing lots. ). contrast? the sa,e studies widel. agreed that CCTV de,onstrated little or no a##ect on reducing cri,es o# passion. Incidents li8e pu;lic drun8enness or acts o# rage generall. did not see, to ;e a##ected ;. the presence o# CCTV ca,eras. This #its a ;roadl. accepted rational actor ,odel and the e##ect that CCTV ca,eras has on rational actors. Since CCTV ca,eras increases the ris8 that a cri,inal will ;e prosecuted #or a cri,e? the cri,inal will respond accordingl.. The ca,eras will a##ect the perceived ris8/reward calculation. Co,,on sense indicates that this i,pact is ,uch ,ore li8el. #or propert./pre,ediated cri,es than it would ;e #or cri,es o# passionate? where ;. de#inition? people are not calculating the conseAuences. !ather than engage in political de;ates over the issue in general? we should use this ,ore nuanced 8nowledge to opti,i2e our use o# CCTV. &in$in"5 CC V as Sin"ular? Sta#le ec%nolo"y The studies overwhel,ing treated CCTV as a singular? sta;le technolog.. &nl. I"Video5ar8et.In#o 14$
the :7 Ho,e &##ice !eport o# 244* even ac8nowledged the issue o# technological change. The rest o# the studies do not even discuss di##erences in technolog. availa;le. I do not #ault the, as the evidence availa;le is li,ited to even conduct such a test. evertheless? di##erences in technolog. can ,a8e an e+tre,e di##erence. The studies cover a ver. ;road ti,e period. The oldest stud. I #ound was #ro, 1$$( with ,ost o# the studies availa;le ;eing per#or,ed in the period #ro, 2444 J 244(. The pro;le, is that CCTV technolog. has e+perienced a dra,atic trans#or,ation in that ti,e period. This is so,ewhat si,ilar to the t.pe o# change e+perienced with ,o;ile phones going #ro, ;ig? ;ul8.? li,ited and e+pensive to sli,? power#ul and u;iAutious. It is Auite un#air to assess the Auestion is CCTV e##ective? in an. #or,? without #actoring in the di##erences in the t.pe o# technolog. used. The e+a,ples #ound in the studies were #airl. shoc8ing co,pared to toda.<s ,ainstrea, CCTV s.ste,s. The clear ,aEorit. o# s.ste,s e,plo.ed in the studies used VC!s. 'ven when s.ste,s used DV!s? ,ost were recording under 2 #ra,es per second. -ll o# the s.ste,s used standard de#inition ca,eras. 3hile none o# the reports discussed the t.pe o# trans,ission s.ste,s? given that al,ost all the tests were #ro, 244( or earlier? it is e+tre,el. li8el. none o# the, were using I" networ8s #or trans,ission. Though li,ited? the ;est discussion on this topic in the literature is the 244* :7 Ho,e &##ice !eport. 3hile none o# this is the researcher<s #ault? not #actoring changes in technolog. o;scures crucial di##erences. In the reco,,endations section? I e+plore what t.pes o# technologies and how the. can i,pact s.ste, e##ectiveness. &in$in"5 Differences in Per Camera Costs (ar"ely I"nore$ 3hile ,ost studies cited general cost nu,;ers? the cost per ca,era was largel. ignored. The ,ost #reAuentl. cited nu,;er is the a,ount the :7 ho,e o##ice has I"Video5ar8et.In#o 114
spent on CCTV =*44 ,illion pounds>. However? onl. the 244* :7 Ho,e &##ice stud. actuall. ;ro8e down the cost per ca,era. Since the studies were #ocused on deter,ining i# the cri,e rate was reduced? this ele,ent is understanda;le. evertheless? co,,unities could save signi#icant ,one. and i,prove e##ectiveness ;. ,ore care#ull. trac8ing the cost per ca,era. :nderstanding the cost per ca,era is i,portant to recogni2e changes in technolog. and to identi#. waste. The 244* :7 Ho,e &##ice report indicated that cost per ca,era ranged #ro, L1?444 pounds to L%%?444 pounds #or ca,eras installed in the late 1$$4s and earl. 2444s. The stud. does not clearl. e+plain the cause o# the cost di##erences. In ,. e+perience deplo.ing si,ilar s.ste,s? the ,ain driver o# costs #ro, this era is the trans,ission s.ste,s. )ecause these ca,eras are generall. outdoors and distri;uted throughout a cit.? trans,ission s.ste,s need to ;e ;uilt to send the video #ro, the ca,era to the ,onitoring center. The solution o# choice in this ti,e #ra,e was proprietar. analog #i;er trans,ission s.ste,s. Such s.ste,s reAuired e+pensive trans,ission eAuip,ent and al,ost alwa.s la.ing o# new #i;er. This routinel. generated costs o# thousands to tens o# thousands. ). contrast? toda.? the solution o# choice #or trans,ission is I" networ8s. I" networ8s dra,aticall. reduce the cost o# trans,ission. I" networ8s replace proprietar. analog #i;er s.ste,s with low cost co,,odit. I" eAuip,ent. I" etwor8s o#ten can share e+isting #i;er networ8s or connect to a teleco,,unication carriers s.ste, to greatl. reduce or eli,inate the need #or new #i;er or construction. &in$in"5 Cops 's Cameras Comparison Counterpro$ucti'e - #reAuent senti,ent e+pressed ;. interviewees in ;oth articles and studies is the pre#erence #or police o##icers versus ca,eras. 3hile this is o;viousl. understanda;le and I e+pect ,ost ever. reasona;le person would agree that police o##icers are pre#era;le to ca,eras? this o,its a crucial ele,ent.
I"Video5ar8et.In#o
111
Since police o##icers are so ,uch ,ore e+pensive than ca,eras? a co,parison ;etween the two is ver. ,isleading. -ccording to the 244* :7 !eport? the annuali2ed cost per ca,era ranged #ro, /44 to %444 pounds. This is 1/1*th to 1/84th the cost #or a .earl. police o##icer =including ;ene#its? training? eAuip,ent? etc>. In actualit.? then the co,parison is ,ore li8e do2ens o# ca,eras versus an o##icer. @urther,ore? given the signi#icant price reduction in CCTV s.ste,s since ,ost o# the tests occurred? the co,parison is now ;etween hundreds o# ca,eras and a single police o##icer. '+a,ined at a ,acro level? a si,ilar distortion is apparent. 5an. o# the articles and studies cite the *44 ,illion pound :7 Ho,e &##ice spending over the last decade. evertheless? *44 ,illion pounds #or CCTV represents less than 1K o# the spending on police o##icers during that ti,e period. 'ven i# all #unding on CCTV was trans#erred to hire new police o##icers? it would onl. increase #unding ;. a ver. s,all percent. -nd? o# course? a s,all percent increase in police o##icers would not ;e e+pected to dra,aticall. decrease cri,e either. This is an area where CCTV proponents have created unrealistic e+pectations that actuall. under,ine their own cause. ow? let<s e+a,ine so,e reco,,endations: Recommen$ation5 A#an$on emp%asis on "eneral crime re$uction "roponents o# pu;lic CCTV s.ste,s should a;andon the e,phasis and clai, that CCTV s.ste,s can reduce cri,e generall.. 'ven i# proponents ignore the #act that studies de,onstrate this? clinging to this clai, onl. creates greater de;ate and dissension. ). a;andoning this clai,? it will heal so,e o# the ,aEor discord and allow all parties to #ocus on ;etter uses o# CCTV. Given the vastl. i,proved Aualit. o# toda.<s CCTV s.ste,s at greatl. reduced prices? this should ;e reasona;le to acco,plish. I"Video5ar8et.In#o 112
Recommen$ation5 &ocus ProCects on Crime Sol'in" Iust li8e the private sector has ;roadl. adopted CCTV ;. #ocusing on solving cri,es? the pu;lic sector should too. This would save co,,unities ,one. as certain #eatures or ca,eras could ;e eli,inated and designs could ;e #ocused on areas and technologies that help solve cri,es. This would si,ultaneousl. ease the i,pact on privac. as less attention and resources would ;e placed on tr.ing to ,onitor s.ste,s live and there;. the ris8 o# ,onitoring the innocent pu;lic. Recommen$ation5 &ocus ProCects on Material>Preme$itati'e Crimes To the e+tent that CCTV is used to support cri,e reduction? such e##orts should #ocus on ,aterial/pre,editative cri,es. 0i,iting the locations covered and ,onitored live to those areas with high rates o# these t.pes o# cri,es will ,a+i,i2e the pro;a;ilit. those s.ste,s will ;e e##ective. Recommen$ation5 ar"et ec%nolo"ies t%at Support Crime Sol'in" Though historicall. the ca,era o# choice has ;een a "TH? s.ste,s should e,phasis the use o# ,egapi+el #i+ed ca,eras. - "TH? or "an/Tilt/Hoo, ca,era? can ;e controlled ;. an operator to loo8 in ,an. di##erent directions and areas. "TH ca,eras are #avored ;. securit. operators as it allows the, to control the ca,era in live ,onitoring. Two signi#icant downsides e+ist #or "THs: &ne? the. reAuire a dedicated operator to use the ca,eras? incurring signi#icant operation cost. Two? "THs are generall. ;ad #or producing evidence ;ecause the. ,iss ever.thing e+pect #or the area where the ca,era is ,o,entaril. positioned. 5egapi+el ca,eras are #ar ;etter #it #or pu;lic places and cri,e solving. The ca,eras ;eing used in the stud. are standard de#inition units with ver. li,ited a;ilities to view details. 5uch li8e the transition #ro, #il, ca,eras to toda.<s I"Video5ar8et.In#o 11%
digital super high resolution ca,eras? CCTV s.ste,s now routinel. e,plo. ,egapi+el ca,eras that provide dra,aticall. greater detail. Such detail is 8e. #or pu;lic places that usuall. cover large outdoor areas. In this scenario? ,egapi+el ca,eras give .ou the ;ene#its o# "TH ca,eras ;ecause the detail allows 2oo,ing with the ;ene#its o# a #i+ed ca,eras< a;ilit. to alwa.s capture video o# a set area. This is critical to cri,e solving ;ecause the ca,era needs to have an i,age i# we are to use the evidence to identi#. or prosecute a cri,e. Recommen$ation5 Minimi1e Cost Per Camera Given what we have learned #ro, #irst generation s.ste,s and the advances in technolog. availa;le toda.? we should ;e vigilant a;out trac8ing and ,ini,i2ing the cost per ca,era. 3e now have a good sense o# what wor8s and does not wor8. 3e should opti,i2e around that and ensure that we can 8eep costs per ca,era low. The two 8e. ele,ents in ,ini,i2ing costs is =1> ensuring that I" networ8s are leveraged and that =2> unnecessar. #unding is not spent #or needless ;ells and whistles. ). doing this? ,unicipalities should easil. ;e a;le to deplo. s.ste,s #or ;etween 2?444 and (?444 pounds per ca,era. This would drop the cost ;. /4K or ,ore relative to historical standards. Conclusion 3ith our e+tensive e+perience and 8nowledge? we ,ust re6position goals? ,odi#. designs and econo,i2e our e##orts:
Set the goals appropriatel. on tas8s that can succeed: Cri,e Solving and "ropert. Cri,e !eduction Select technologies such as I" and ,egapi+el ca,eras that i,prove per#or,ance 'nsure spending per ca,era is controlled and ;ene#its #ro, new technologies 11(
I"Video5ar8et.In#o
3ith these practices? we can ensure ;oth e##ective CCTV s.ste,s and a positive econo,ic contri;ution to societ..
I"Video5ar8et.In#o
11*
III Evaluating New Products
I"Video5ar8et.In#o
11/
Chapter 2$:
How to Read 4ar6eting 4aterial
-l,ost all I" video in#o is vendor ,ar8eting. Good decision ,a8ing reAuires criticall. reading and anal.2ing this ,aterial. -t #irst? I did not ;elieve that ,ost in#or,ation was vendor ,ar8eting ,aterial. &;viousl.? we; sites and press releases are ,ar8eting ,aterials ;ut .ou also have articles and reports #ro, ,aga2ines. However? al,ost ever. article I #ind across a do2en ,aga2ines is written ;. a vendor =usuall. the head o# ,ar8eting>. 5oreover? ,ost o# those articles are clearl. pro,otion pieces #or the vendor<s o##erings. The. argue the ,erits o# the trends ;ehind their co,pan.<s o##erings with ,ini,al attention or #air treat,ent o# opposing views. 'ven news reports are routinel. copies or e+cerpts o# press releases. -s such? .ou reall. need to ;e care#ul and cogni2ant o# the ,otivation and structure o# the in#or,ation .ou are reading. I have had to re6train ,.sel# to ;e ,ore critical o# what I read as I reali2e how consistentl. this is an issue. I# .ou want to ,a8e good decisions and Auic8l. discern what is the true value o# what .ou are reading? I encourage .ou tr. the techniAues I share here. )etter anal.sis o# this in#or,ation can reall. save .ou #ro, ,ista8es or #uture pro;le,s. -t the sa,e ti,e? I a, hoping vendor<s consider ,odi#.ing their ,ar8eting ,aterials. -s I will discuss throughout? in the long run? I ;elieve all parties will ;ene#it #ro, clearer co,,unication.
Here are ,. 8e. reco,,endations #or reading ,ar8eting ,aterial: I"Video5ar8et.In#o 111
How to !ead 5ar8eting 5aterial
1. Deter,ine how well the o##ering wor8s 2. Deter,ine what ;ene#its the o##ering provides over the ne+t ;est alternative %. Deter,ine what the cost o# the o##ering is
4D +ow well it wor*s 5ar8eting ,aterial routinel. spea8 in glowing ter,s o# what their o##ering does. This is great #or esta;lishing the conceptual potential o# a product? which is a necessar. ele,ent o# co,,unicating value. It sets the stage #or what is #unda,entall. di##erent and what custo,ers ,ight e+pect to gain #ro, the o##ering. The pro;le, is that it is so vague that it is i,possi;le #or readers to deter,ine how well it #its #or their environ,ent. 5ost i,portantl.? ver. rarel. does the ,aterial discuss how well the o##ering wor8s or how well it ,ight wor8 in di##erent applications. I have seen this happen #or 2 reasons: =1> the vendor is not sure which seg,ent the product is a #it or =2> the vendor wants to launch the widest possi;le net and not lose an. prospects. In either scenario? it ;eco,es ver. hard #or a reader to ,a8e a realistic deter,ination o# the #it #or their needs. I do not thin8 this is ulti,atel. ;ene#icial #or an. o# the parties. The vendor ,ight get a short ter, win ;. an i,,ediate sale. However? even #or the vendor? it still could ;e a pro;le,. I# the deplo.,ent goes poorl. =and o#ten does i# the #it is poor>? the chances #or repeat ;usiness and re#errals is low. 'ssentiall. it ;eco,es a ver. high cost sale that does not grow the long ter, ,ar8et. -s a reader? .ou need to clearl. as8 .oursel# how well this o##ering will wor8. Consider what operational or environ,ental issues ,a. under,ine the proEect. Since it is unli8el. .ou will get a clear and #air assess,ent #ro, a vendor? .ou
I"Video5ar8et.In#o
118
need to do this .oursel# to ,a8e good decisions.
7D %e ne0t #est alternati'e 5ost ,ar8eting ,aterial gloss over the ;ene#its o# .our e+isting s.ste,s or processes. @or instance? V! vendors routinel. clai, ;ene#its that an. low end DV! can deliver. 5egapi+el vendors ,a8e assu,ptions a;out ca,era deplo.,ents that .ou would al,ost never use in deplo.,ent. 'ssentiall.? the co,parisons are s8ewed to ,a+i,i2e the positive positioning o# their products. = ote: this is not uniAue to an. product categor.? V!s and ,egapi+el ca,eras are si,pl. two o# the ;ig products o# the da.>. This causes con#usion a;out the speci#ic di##erentiators o# the product o##ering. Trul. innovative aspects can ;e lost in long lists o# routine e+isting #eatures and #unctionalities. 'nd users can ;e ,otivated to purchase ,ore co,ple+ or e+pensive products that do not trul. generate ,ore value #or their organi2ations. 3hile it is hard #or vendors to trul. understand co,petitor<s o##erings deepl.? ,ore clearl. and #airl. stating actual advantages can help custo,ers ,a8e ;etter decisions ,ore Auic8l.. Though I honestl. have little hope o# this ele,ent changing? clearl. considering what trul. is a new ;ene#it can help deter,ine the actual value #or .our organi2ation.
8D %e cost of t%e offerin" Vendors rarel. discuss costs o# their o##ering. Generall.? vague state,ents are o##ered li8e <su;stantial !&I< or <signi#icantl. increased value.< Vendors are Eusti#ia;l. concerned a;out inter#ering with their dealer<s a;ilit. to set end user pricing. The. are also o#ten worried that disclosing price will scare o## so,e
I"Video5ar8et.In#o
11$
;u.ers and that it is ;etter to pro,ote their general ;ene#its and handle pricing once the custo,er is engaged. The huge downside o# not discussing costs is that it<s i,possi;le #or readers to deter,ine <value< or <!&I<. 3ithout having an idea o# cost? ;. de#inition? .ou cannot calculate #inancial return. -nd it<s not Eust a ,athe,atical issue. This is a ver. practical issue as readers cannot discern whether an o##ering is #easi;le #or their ;udgets. I see this all the ti,e with articles on !-ID? QoS? I" ,ulticast? redundant servers. The costs #or these #eatures/products can ;e ver. e+pensive. It is hard #or an.one to assess #it without having a ;allpar8 sense o# cost. It would ;e ver. valua;le i# vendors provided rough costs #or their products. It does not need to ;e a negotiated price? a si,ple 5S!" would wor8 #ine. !eaders need to 8now the general range pricing is in. @or instance? is .our ,egapi+el ca,era close to L*44? L1444? L1*44? L24449 Setting an appro+i,ate range is good enough to allow a reader to gauge how that would #it in their ;udgets and how ,uch value the product would need to deliver.
7eeping these points in ,. ,ind when .ou read ,ar8eting ,aterial can help .ou ;etter assess the true value o# the o##erings ;eing pro,oted. :ntil ,ar8eting ,aterials ;eco,e ,ore clear =i# ever>? appl.ing this should help in evaluating this in#or,ation.
Chapter 2&:
How to #valuate New -echnolog*
5ost new technolog. #ails ;ut when it is success#ul? the ;usiness ;ene#its can ;e I"Video5ar8et.In#o 124
How to 'valuate ew Technolog.
enor,ous. The challenge then is how to e##icientl. deter,ine what new technolog. is legit so that .ou si,ultaneousl. avoid disaster and reap the rewards o# the rare ge,. Fou ,a. have do2ens o# co,panies to review. 'ach new pro,ising technolog. spurs the entrance o# ,an. co,panies hoping to ena;le the technolog.. So it<s not Eust evaluating the technolog.? it<s #iguring out which co,panies? i# an.? has the winning solution. Fou usuall. cannot ,a8e the evaluation ;ased purel. on .our own 8nowledge. 5ost o# the ti,e when .ou are evaluating a new technolog.? .ou lac8 speci#ic technical e+pertise in that area. -s such? .ou need to #igure out tactics and techniAues to give .oursel# the ;est chance o# proEecting winners. This article e+plores * 8e. tips I have learned over the .ears wor8ing as an integrator and ,anu#acturer. Here the. are: 1. Veri#. 5ar8eting 5aterials "rovide Technical Details 2. -s8 Speci#ic Questions -;out "ro;le,s with the "roduct %. Veri#. that the Vendor is not a "athological 0iar (. -s8 the Vendor how the product will wor8 with all ele,ents o# .our operations *. Test :nder Stress
Does t%e Mar*etin" Materials pro'i$e ec%nical Details, The ver. #irst thing .ou should do is chec8 how technical the ,ar8eting ,aterials I"Video5ar8et.In#o 121
are. Fou do not need to 8now the technical Eargon. -t #irst? si,pl. scan and notice how ,uch o# the ,ar8eting ,aterials are prose =li8e an essa.> versus how ,uch are acron.,s? nu,;ers? diagra,s? etc. @ew technical details are a strong indicator that the product is either conceptual or vaporware. &#ten? the lac8 o# technical details arises ;ecause the co,pan. is pro,oting an idea ;ut the. are wea8 in engineering. &ther ti,es? their engineering is #ine ;ut the product is still so earl. that the. have not gotten #ar enough to #igure out a lot o# the technical details. I generall. discard co,panies #ro, #urther consideration that do not ,eet this criteria. &n the other hand? Eust ;ecause a co,pan. does have technical details? does not ,ean it will de#initel. wor8. The co,pan. ,a. ;e especiall. sophisticated in ,ar8eting or there ,a. ;e ,ore issues. -s such? si,pl. treat this as a #irst gate.
Are you as*in" Specific Euestions a#out Pro#lems, 5ost people will not lie to .ou ;ut are &7 with not telling .ou the truth. Since people are generall. unco,#orta;le l.ing? a co,,on tactic is to ignore discussing da,aging issues. I# .ou as8 a vendor BHow ,an. co,panies are using .our product in production9B? ,ost vendors will tell so,ething close to the truth. I# .ou do not as8 an.thing? al,ost no one will volunteer that the product has never ;een deplo.ed or onl. deplo.ed at 1 or 2 sites. Strictl. spea8ing? the. are not l.ing to .ou ;ut the outco,e is si,ilar ;ecause it leads .ou to ;elieve incorrectl. a;out a 8e. ele,ent in the decision ,a8ing process. :nli8e ,ature products where it is reasona;le to ta8e things #or granted? this is a great ris8 with new technolog. products. The challenge is new technolog. products alwa.s have pro;le,s. That does not
I"Video5ar8et.In#o
122
,ean .ou should not use the, ;ut .ou have to ;e aware o# what those pro;le,s are. )e e+plicit and as8 things li8e: How ,an. sites have the product deplo.ed9 3hat was the cause o# the last % #ailures o# .our product in the #ield9 3hat was the cause o# the product #ailing in previous pilots9 =all products #ail in at least so,e pilots> Can I have a re#erence9 =Do not accept the e+cuse that the. cannot tell .ou ;ecause o# securit. issues. -n. product with success has at least a #ew custo,ers willing to tal8? especiall. i# .ou are a securit. ,anager.>
Iust re,e,;er? do not ta8es things #or granted? ,a8e sure to as8.
Is t%e Ven$or a Pat%olo"ical (iar, "athological liars are a ver. dangerous #orce in new technolog. products. 'ver. once in a while? a vendor will consistentl. spin and de#lect an. pro;le,s or criticis,s. The. will ;e so good that .ou will rela+ .our guard and in .our enthusias, #or the ;ene#its o# the pro;le, will overloo8 pro;le,s. This is dou;l. dangerous. @irst? this under,ines .our due diligence ;ut? secondl.? and ,uch worse? pathological liars usuall. have worse products ;ecause the. are too ;us. spinning rather than ;uilding. I e+perienced this when I was an integrator. 3e would go into ,eetings and this gu. would consistentl. spin our o##erings? de#lecting an. legiti,ate issues and creating the perception o# no ris8 and huge reward. &ne ti,e? a custo,er as8ed a technical Auestion li8e BDo .ou use "rotocol O9B and this gu. shot ;ac8 B&# course.B The custo,er? who was #airl. technical? and ,.sel# were ;oth ta8en a;ac8. :n#ortunatel.? what ,. colleague did not understand was that this was an I"Video5ar8et.In#o 12%
outdated protocol that no one wanted to use an.,ore. 3hen we le#t the ,eeting I as8ed hi, wh. he said that. His response was? BI was tr.ing to tell the, what the. wanted to hear.B 5a8e sure vendors are not si,pl. telling .ou what the. want to hear. The ;est tactic to handle this is to as8 another person at the vendor =usuall. a technical person> Auestions awa. #ro, the potential liar. ow ,ost people 8now whether there colleagues are liars ;ut the. are going to ;e Auite reluctant to sa. it directl.. Tal8 to the, a;out operational issues and as8 this person direct Auestions. Fou will get a good sense o# issues and discrepancies Auite Auic8l. this wa..
+ow Does it affect t%e Elements of ;our Operation, ew technolog. products usuall. #ail ;ecause o# un#oreseen operational issues. Generall. it is #airl. eas. to #igure out i# the technolog. solves a ;usiness pro;le,. &n the other hand? it is ver. hard to deter,ine what the operational issues .ou ,ight have deplo.ing and using this technolog.. This is the ,ost i,portant step in evaluating new technolog. products. !egardless o# whatever has ;een said or pro,ised? regardless o# the potential? how the technolog. i,pacts .our operations ,a8es or ;rea8s its via;ilit.. Ver. o#ten? the technolog. results in hidden increases in cost or can si,pl. not ;e ,ade to wor8 with .our e+isting s.ste,s or procedures. Fou ,ust ,a8e sure .ou understand how new technolog. interacts with e+isting s.ste,s. Fou have e+isting s.ste,s and .ou want those s.ste,s to continue to wor8. Fou o#ten #ind out that this technolog. does not wor8 with a 8e. co,ponent o# .our e+isting s.ste,. -s an integrator? I once had a ,aEor pro;le, designing a video anal.tic s.ste, ;ecause it did not integrate with the custo,er<s e+isting
I"Video5ar8et.In#o
12(
,atri+ switch. - ,inor technical detail ;ut it was a ver. serious operational issue. @or all aspect o# .our s.ste,? go through the, and ,a8e sure that there are no hidden operational inco,pati;ilities. Si,ilarl.? while it is eas. to deter,ine the direct cost o# the new technolog. product? .ou ,ust ;e care#ul a;out indirect costs this product ,ight result in. &#ten new technologies will have reAuire,ents that cannot easil. ;e ,et with .our operations. This technolog. ,ight reAuire ,uch greater a,ounts o# ;andwidth or client "Cs that are ,uch ,ore power#ul than .our e+isting ones or signi#icant a,ounts o# training or ,aintenance. 3hen .ou are esti,ating .our costs? ;e sure to consider what the indirect costs can ;e 6 the. o#ten turn a pro,ising proEect into an unrealistic one. The technolog. ,a. ;e good ;ut not good enough #or .our ;usiness o;Eectives. Fou have to ;e sure that it is trul. good enough or .ou will cause a serious operational pro;le,. &#ten? technolog. e+ists to auto,ate e+isting processes ,anaged ;. people. It is Auite co,,on that new technolog. can do a Eo; $4K ot $*K as good as a person. However? in ,an. situations? #ro, an operational or custo,er support standpoint? sacri#icing that *K or 14K can ;e a signi#icant ;usiness pro;le,. I# .ou use #acial recognition to veri#. a person co,ing through a door =auto,ating access control guard veri#ication>? i# that #acial recognition s.ste, ,a8es a ,ista8e onl. *K o# the ti,e? that can ;e * to 24 people a da. that are #rustrated. This ,ight ;e a ver. good s.ste, and strong technolog. ;ut it ,a. not ;e good enough to ,eet the other ;usiness o;Eectives or .our organi2ation. I# .ou do a care#ul assess,ent o# s.ste, interopera;ilit.? indirect costs and con#or,ance with ;usiness o;Eectives and it passes? .ou are ver. li8el. to have a winner.
+ow Does it wor* un$er Stress,
I"Video5ar8et.In#o
12*
&ne 8e. wa. to deter,ine how the new technolog. product a##ects .our operations is ;. doing a pilot. "ilots are co,,on so I onl. have 2 pieces o# advice here. &ne? ,a8e sure .our pilot places the s.ste, under the highest level o# stress .ou e+pect the product to ;e used at in production. &#ten? the test is done in a la; or in .our o##ice. This is a ver. ;ad idea. &##ice or la; test hide issues and wor8s to the advantage o# unscrupulous vendors. How capa;le a product is to handling e+tre,e conditions and loads is a ver. co,,on di##erence ;etween new and ,ature products. It ta8es a lot o# ti,e and e+perience #or a product to incorporate real world challenges and ;e opti,i2ed #or per#or,ance in e+tre,e conditions. "lacing the product in .our toughest operational environ,ent is the ;est wa. to show how read. the product is #or production use. This wa.? an. shortco,ings are e+posed Auic8l. rather than ,onths later a#ter the proEect is well under wa. and it is ver. hard to adEust.
:sing new technolog. products is the ,ost power#ul wa. to generate a ;usiness advantage. I# .ou are a securit. ,anager? it can ena;le .ou to trul. standout and advance in .our career. I# .ou are an integrator? it can drive incredi;le growth. I a, a ;ig proponent o# using new technolog. products. 5a8ing the right decisions a;out new technolog. products is critical. Consider using these steps and hope#ull. .ou will ;e a;le to ,a8e ;etter decisions in less ti,e.
I"Video5ar8et.In#o
12/
How to Calculate Video Surveillance !&Is
Chapter 2+:
How to Calculate Video Surveillance R"Is
!&I calculations are power#ul ;ut can ;e distorted. 3hile the. hold the pro,ise o# identi#.ing o;Eective value? the. can o#ten o;scure the truth. The goal o# this review is to help the securit. ,anager ;etter understand supplier !&I calculations and allow the ,anager to ,odi#. or adEust #or accurate and realistic results. Integrators and ,anu#acturers could also ;ene#it #ro, appl.ing these principles. Good !&I calculation reAuire understanding operational details ,ore than the. do ,ath or ,one.. &nce .ou understand the operational details? the ,ath and ,one. are si,ple. Here are the ( principles in preparing a !&I calculation:

:nderstand the alternative to this proposed invest,ent :nderstand the #ull cost :nderstand the technological de#iciencies o# this invest,ent Veri#. that operational assu,ptions are correct
Principle 345 Alternati'es The ,ost ;asic tric8 to pla. in !&I anal.sis is to choose an alternative that is clearl. ;ad ;ut not relevant to .our case. 5ost vendor !&Is do this. &ne topical e+a,ple is with V!s. @reAuentl.? V!s ,a8e clai,s that the. drive !&I ;. ena;ling centrali2ed ,onitoring or integrating with applications li8e "&S or access control. 3hile certainl. true? #ro, an !&I perspective? this is irrelevant ;ecause DV!s do the sa,e things. It does not ,a8e sense #or a securit. ,anager I"Video5ar8et.In#o 121
to co,pare an V! to a VC! or to nothing ;ecause al,ost ever.one has a DV! or would consider a DV! as an alternative to a V!. To ,a8e a ;usiness case #or the V!? it needs to ;e co,pared to a DV!. @or instance? i# an V! cost BL14?444B and a DV! cost BL8?444B? the invest,ent #or purpose o# calculating the !&I would ;e L2?444 =the pre,iu, #or the V! over the DV!>. -t the sa,e ti,e? the V! could onl. clai, returns on a;ilities that it uniAuel. has over the DV!? there;. eli,inating #ro, consideration aspects such as centrali2ed ,onitoring and application integration. I# .ou do not ta8e this approach and si,pl. calculate an !&I o# an V! versus a VC!? .ou could ;e wasting ,one. ;. pa.ing e+tra #or a V! when a DV! could have delivered the sa,e value. &T': I thin8 V!s o#ten generate ,ore value than DV!s so this is not a criticis, o# V!s. This is a critiAue o# the process o#ten used to Eusti#. V! purchasing decisions. 5egapi+el ca,era suppliers o#ten advocate ca,era eli,ination ;ut this can so,eti,es distort !&I calculations. @or instance? a recent whitepaper e+a,ined a scenario where 1% analog ca,eras could ;e replaced ;. =2> % 5egapi+el ca,eras #or covering a 144 #oot wide outdoor area. The paper concluded that the ,egapi+el ca,era solution was actuall. cheaper. This assu,ption is ,isleading ;ecause the alternative here is reall. using 2 or % analog ca,eras. That is what ,ost securit. ,anagers use toda. and with that as the alternative the cost o# the ,egapi+el ca,era scenario is signi#icantl. higher than analog ca,eras. &T': The ,egapi+el ca,eras in this scenario ,a. deliver ,uch higher !&I ;. ;eing a;le to solve previousl. unsolva;le cases due to their greater Aualit.. I a, not o;Eecting to the design? si,pl. the ,ethod on how the #inancial Eusti#ication was ;eing ,ade.
I"Video5ar8et.In#o
128
The securit. ,anager and ,egapi+el vendor should concentrate on de,onstrating the increased return delivered speci#icall. ;. the enhanced i,age Aualit.. Speci#icall.? onl. cases solved with a ,egapi+el ca,era that could not ;e solved ;. an alternative analog ca,era should ;e #actored in the !&I #or ,egapi+el ca,eras. I# identi#.ing a license plate was critical in solving a a case? the ,egapi+el ca,era should get credit #or it. )ut i# the case could ;e solved ;. identi#.ing that the car was a white Civic? an analog ca,era would ;e eAuall. capa;le and the ,egapi+el ca,era should not get credit. This distinction is routinel. ;lurred ;ut i# .ou are to trul. deter,ine an accurate !&I? this is a critical #actor.
Principle 375 =n$erstan$ t%e full cost &#ten? vendor supplied !&Is leave out indirect costs. These ;eco,e hidden costs that can drag .our true !&I down signi#icantl.. &ne o# the hidden costs o# video anal.tics is the need #or ,onitoring. Depending on the level o# #alse alerts? .ou ,a. need to dedicate resources to assess and veri#. the alerts. This cost could ;eco,e Auite signi#icant. Fou ,a. ;e a;le to get the technolog. to wor8 as advertise ;ut .ou ,a. need to dedicate e+tra operational resources to ;ring it to that level. 5a8e sure .ou understand what i# an. indirect costs are needed and #actor this in. 5egapi+el ca,eras are another e+a,ple o# indirect costs. 3ith ,egapi+el ca,eras? it is not onl. the increased ca,era cost ;ut the increased cost o# the storage and ;andwidth. -l,ost all ,egapi+el ca,eras in production use ,uch ,ore ine##icient co,pression than analog ca,eras. -lso? i# .ou trul. want enhanced resolution in ,egapi+el ca,eras? this will #urther increase storage costs =and o#ten networ8 costs>.
I"Video5ar8et.In#o
12$
-gain? these ;oth ,a. ;e Eusti#ia;le ;ut a #air anal.sis ,ost include an. additional cost #or the,.
Principle 385 ec%nolo"ical Deficiencies 3hen a vendor provides .ou an !&I? usuall. it assu,es that the technolog. wor8s as advertised. 3ith new technolog. that so,eti,es turns out not to ;e the case. -lso? so,eti,es? the technolog. wor8s ;ut not in the circu,stances .ou need it in. This is one o# the 8e. issues with video anal.tics. It is eas. to sa. that peri,eter violation has the potential to reduce losses signi#icantl.. However? it depends on how well it wor8s. I# it turns out that .our #acilities have a lot o# snow? the s.ste, ,a. not wor8 properl. during those ti,es. This can reduce the potential loss reduction proEected. Si,ilarl.? .ou ,a. want to use a ,egapi+el ca,era to capture license plates and #aces in a ver. dar8 area at night. 5an. ,egapi+el ca,eras wor8 poorl. with low light conditions. I# .ou were proEecting to solve cases during this ti,e? this ,a. not actuall. wor8. Si,ilarl.? the s.ste, ,a. turn out to ;e too hard to use so that .our operators #ail to solve as ,an. cases as the technolog. ,ight potentiall. deliver. Care#ull. review what the vendor<s proEections are and ,a8e sure that an. technological de#iciencies are re#lected in the !&I calculation.
Principle 395 Operational Assumptions Suppliers can onl. ,a8e ;est guesses as to the operational realities o# a securit.
I"Video5ar8et.In#o
1%4
,anager. &#ten those guesses are ver. opti,istic or si,pl. do not ,atch .our organi2ation<s situation. '+a,ples o# these assu,ptions include loss per incident? nu,;er o# incidents per ,onth? nu,;er o# incidences that this s.ste, will solve. @irst? .ou need to as8 and understand what these operational assu,ptions are in a vendor provided !&I. Co,pare that to .our actual ,etrics and re6adEust to deter,ine appropriate levels. How ,uch ti,e does the s.ste, reall. save .ou9 How ,an. incidents per .ear can .ou reall. solve with the new s.ste, that .ou could not with old9 It<s pro;a;l. going to di##er #ro, the vendor assu,ptions? so ;e read. to adEust the !&I calculations.
The challenge in all #inancial ,odels is the assu,ptions ,ade. ). using these ( principles? .ou can ;etter assess and deter,ine the right assu,ptions to ,a8e. Identi#. hidden costs and pro;le,s that a theoretical !&I ,a. ignore and 8eep .our suppliers honest. :ntangle co,,on !&I con#usions and distortions and .ou will ;e rewarded with an accurate !&I providing clarit. on genuine ;usiness value.
I"Video5ar8et.In#o
1%1
Thank You.
For more information, contact: John Honovich (646) 867-1965 jhonovich@ipvideomarket.info IPVideoMarket.Info

Security Manager Guide Video Surveillance v2 0

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Security Manager Guide Video Surveillance v2 0

Hochgeladen von

Copyright:

Verfügbare Formate

Security Managers Guide to Video Surveillance

Version 2.0 / November 2008 John Honovich IPVideoMarket.Info

About IP Video Market Info

2009 Industry Guide

About the Author

I Introduction to Video Surveillance

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

$. %ideo Management S"stem

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

How to Design Video Surveillance Solutions

Introduction to V!s / I" Video So#tware

Introduction to NVRs / IP Video So tware

NVRs are Software Only Applications (DEBA ED! I"Video5ar8et.In#o 18

Introduction to V!s / I" Video So#tware

NVRs Support Analo" Cameras #y Enco$ers

Introduction to V!s / I" Video So#tware

!ecord Video View 0ive Video 24

Introduction to V!s / I" Video So#tware

Search #or !ecorded Video View !ecorded Video

Conduct these #unctionalities #ro, a re,ote co,puter

&ptions #or Di##erent Verticals/-pplications =!etail? )an8ing? "eri,eter 21

Introduction to V!s / I" Video So#tware

Introduction to V!s / I" Video So#tware

Introduction to Video C"D#Cs

Introduction to Video C&D'Cs

Co,press the individual i,age ;. itsel#

Introduction to Video C&D'Cs

Co,press a series o# i,ages together

Introduction to Video C&D'Cs

Introduction to Video C&D'Cs

Introduction to Video C&D'Cs

%andwidth %asics or Video Surveillance

)andwidth )asics #or Video Surveillance

)andwidth )asics #or Video Surveillance

)andwidth )asics #or Video Surveillance

.Ban$wi$t% Cost/ I" ca,eras 5egapi+el ca,eras 1 5;/s * 5;/s to 14 5;/s

)andwidth )asics #or Video Surveillance

)andwidth )asics #or Video Surveillance

#'a(ining Video )nal*tics

'+a,ining Video -nal.tics

'+a,ining Video -nal.tics

'+a,ining Video -nal.tics

'+a,ining Video -nal.tics

,icense Plate Recognition -utorial

0icense "late !ecognition Tutorial

0icense "late !ecognition Tutorial

0icense "late !ecognition Tutorial

0icense "late !ecognition Tutorial

Introduction to DVR/NVR Storage "pti(i/ation

+ow Do I Optimi1e Stora"e,

Introduction to DV!/ V! Storage &pti,i2ation

Introduction to DV!/ V! Storage &pti,i2ation

Introduction to DV!/ V! Storage &pti,i2ation

Introduction to DV!/ V! Storage &pti,i2ation

Introduction to DV!/ V! Storage &pti,i2ation