Sie sind auf Seite 1von 2

Fusing Sensors for Occupancy Sensing

in Smart Buildings

Nabeel Nasir, Kartik Palani, Amandeep Chugh, Vivek Chil Prakash,


Uddhav Arote, Anand P. Krishnan, and Krithi Ramamritham

Department of Computer Science


Indian Institute of Technology Bombay,
Mumbai, India
{nabeel12,kartik,amandeepchugh12,
vivekcp,uddhava,anandkp,krithi}@cse.iitb.ac.in
http://www.cse.iitb.ac.in

Abstract. Understanding occupant-building interactions helps in per-


sonalized energy and comfort management. However, occupant identi-
cation using aordable infrastructure, remains unresolved. Our analysis
of existing solutions revealed that for a building to have real-time view of
occupancy state and use it intelligently, there needs to be a smart fusion
of aordable, not-necessarily-smart, yet accurate enough sensors. Such a
sensor fusion should aim for minimalistic user intervention while provid-
ing accurate building occupancy data. We describe an occupant detection
system that accurately monitors the occupants count and identities in a
shared oce space, which can be scaled up for a building. Incorporating
aspects from data analytics and sensor fusion with intuition, we have
built a Smart-Door using inexpensive sensors to tackle this problem. It
is a scalable, plug-and-play software architecture for exibly realizing
smart-doors using dierent sensors to monitor buildings with varied oc-
cupancy proles. Further, we show various smart-energy applications of
this occupancy information: detecting anomalous device behaviour and
load forecasting of plug-level loads.

Keywords: Smart Door, Smart Building, Energy Saving, User Comfort,


Electrical Energy.

1 Introduction

Designing new green buildings and retrotting existing buildings with green
technologies pose numerous research challenges but essential for society. Two of
the main motivations for this transition towards a smarter and greener electricity
grid have been capping total usage or attening the peak and reducing the
carbon footprint and costs. This has sparked new interest in developing smarter

The authors would like to thank DeitY, Govt. of India and TCS for their generous
support of this work.

Corresponding author.

R. Natarajan et al. (Eds.): ICDCIT 2015, LNCS 8956, pp. 7392, 2015.

c Springer International Publishing Switzerland 2015
74 N. Nasir et al.

Table 1. Approaches to tracking occupancy using various sensors and their fusion

Occupancy
Sensor Advantages Disadvantages
Information

Cheap;
When users become station-
Passive Scalable; RT Re- Presence of oc-
ary (eg., working on PC) room
Infra Red sponse; No cupants in room
occupancy detected as NIL
User Intervention

Cheap; User inter- Response is not real-time; Ac-


CO2 [3] vention is not re- curacy reduces when there is Presence of oc-
quired; Scalable proper air circulation cupants in room

Accurate when User should carry RFID tag;


Radio
proper measures Tags must not be kept near
Frequency Count and iden-
taken;Real Time metallic objects; Accuracy de-
Identication tity
response; No user pends on speed of walking[5]
intervention req.

Computationally challenging;
Expensive; Accuracy is less
Face No user Interven- for moving objects; Not eas- Count and iden-
Recognition
tion required ily scalable; Requires 2 cam- tity
[9]
eras to detect Entry and Exit

Cheap; Real-Time Not suitable for environments Presence of oc-


Sound Response; like labs and libraries cupant in room
Detection [10] Scalable

The sensor fails to detect oc-


Presence of oc-
PIR+Reed[1] Cheap; Scalable cupancy in a multi user envi-
cupant in room
ronment

Accuracy reduces when users


dont comply to the rules of
WiFi+Lan+ the system, WiFi cant distin- Count and iden-
Cheap;Scalable
IM+Calender guish a person who is right tity
+Access outside the room, will be de-
Badge[6]
tected as inside the room

Das könnte Ihnen auch gefallen