Barrister Ranjit Mohanty International Institution of Technology

Project Work Report

On

LINE FOLLOWING ROBOT

Submitted in partial fulfillment of Live Project

By
PALLAV KUMAR NAYAK [08EE-093]
Under the guidance of

RAJENDRA KUMAR BHOI

Dept. of CS, BRM IIT

Department of Electrical and Electronics Engineering

BRM IIT, PANDRA, BHUBANESWAR
2008-2012

Barrister Ranjit Mohanty International Institution of Technology

River View Campus, Barrister Ranjit Mohanty Marg, Rasulgarh,
Bhubaneswar, Orissa

DEPARTMENT OF ELECTRICAL AND ELECTRONICS

CERTIFICATE
It is certified that the project work entitled LINE FOLLOWING ROBOT carried out by
Mr. Pallav Kumar Nayak [08EE-093] a bonafide student of 3rd Semester in partial
fulfillment for the Live Project of BRM IIT, Bhubaneswar, during the year 2009. It is
certified that the corrections/suggestions indicated for Internal Assessment have been
incorporated in the Report department library. The project report has been approved as it
satisfies the academic requirements in respect of project work prescribed for the said
purpose.

Signature of the Guide

(R.K.BHOI)

Signature of the Principal

(-------------------)

External Viva
Name of the Examiners
1.
2.

Signature with date

ACKNOWLEDGEMENT

An endeavor over long period can be successful only with advice and guidance
of many well wishers. My sincere thanks to the management and -----------,
DEAN, BRM IIT, for providing me the opportunity to conduct my project
works. I am highly indebted to Mr. R.K.BHOI, CS DEPARTMENT, BRM IIT
for his assistance and constant source of encouragement.
I wish to express my profound and deep sense of gratitude to Mr. SASMIT
HOTA, IT DEPARTMENT, BRM IIT for sparing his valuable time to extend
help in every step of my project work.
Last but not the least we would like to thank our friends and family for their
help in every way for the success of this project report.

PALLAV KUMAR NAYAK

CONTENTS
PAGE
1. SYNOPSIS.
2. PREAMBLE...
2.1 INTRODUCTION....
2.2 PROBLEM DEFINITION.....
2.3 OBJECTIVES OF THE STUDY.
2.4 SCOPE OF STUDY.
2.5 REVIEW OF LITERATURE..
2.6 APPLICATIONS.........
2.7 LIMITATIONS....
2.8 METHODOLOGY...
3. THEORY....
3.1 THE DIFFRENTIAL STEERING SYSTEM..
3.2 D.C. MOTORS
3.3 H-BRIDGE MOTOR CONTROL...
3.4 INTELLIGENCE.
3.5 THE AVR-ATMEGA16 MICROCONTROLLER..
3.6 PWM SPEED CONTROL...
3.7 THE PICMICRO CCP MODULES.
4. DESIGN AND IMPLEMENTATION.
4.1

SCHEMATIC....

4.2

PROCESS EXPLANATION.....

4.3

FLOW CHART.....

4.4

CODE....

4.5

CODE EXPLANATION...

5. RESULT & CONCLUSION..

6. BIBLIOGRAPHY..

7. APPENDIX

1. SYNOPSIS
The line following robot, operates as the name specifies. It is programmed to
follow a dark line on a white background and detect turns or deviations and
modify the motors appropriately. The optical sensor is an array of commercially
available IR reflective type sensors.
The core of the robot is the ATMEGA 16 microcontroller. The speed control of
the motors is achieved by the two PWM modules in the C. The direction
control is provided by 2 I/O pins. The H-Bridge motor driving/control chip
takes these signals and translates it into current direction entering the motor
armature. The motors require separate supply for operation.
The differential steering system is used to turn the robot. In this system, each
back wheel has a dedicated motor while the front wheels are free to rotate. To
move in a straight line, both the motors are given the same voltage (same
polarity). To manage a turn of different sharpness, the motor on the side of the
turn required is given lesser voltage. To take a sharp turn, its polarity is
reversed.
The robot uses IR sensors to sense the line. An array of 8 IR LEDs (Tx) and
sensors (Rx), facing the ground has been used in this setup. The output of the
sensors is an analog signal which depends on the amount of light reflected back,
this analog signal is given to the comparator to produce 0s and 1s which are
then fed to the microcontroller.
The control has 6 modes of operation, turn left/right, move left/right, and drift
left/right. The actual action is caused by controlling the direction/speed of the
two motors (the two back wheels), thus causing a turn. The actual
implementation is a behavior based (neural) control with the sensors providing
the inputs. The robot can also be programmed to find the line by pseudo-random

movement in case no line is detected by the optical sensor.

Figure 1.1: Block Diagram of the Line Following Robot

2. PREAMBLE
2.1. INTRODUCTION

The robots of the movies, such as C-3PO and the Terminator are portrayed as
fantastic, intelligent, even dangerous forms of artificial life. However, robots of
today are not exactly the walking, talking intelligent machines of movies,
stories and our dreams.
In the 1970s scientists proposed that in the year 2000 we would have created
artificial life forms, almost perfect in terms of intelligence and capabilities. The
dream of free and efficient labor made the researchers of the time go on day and
night to bring the dream to existence. But the task was futile due to the lack of
compact processors to carry out the calculations which were oh so necessary.
Now in the year 2000, the micro-processor technology is thousands of times
more advanced than what existed back then. But still the robots of today are no
way close to what our movies portray them to be. This is not only due to
drawbacks in processor technology, but also in various other fields such as
vision, motor control so and so forth.
Robots may never make it to our kitchens or living rooms as personal slaves,
but they certainly have made their way to the manufacturing industry, aerospace industry, and yes to the work benches of robotic hobbyists. Robots are
now working in dangerous places, such as nuclear disposal, space explorers, fire
fighting and many more.
The word "robot" originates from the Czech word for forced labor or serf.
Robots are electronic devices intended to perform a desired function. Many
refer to them as "machines", however, a drill press is a machine, yet it requires
an operator to perform its function, where robots can be programmed to do it

themselves. Robots have the potential to change our economy, our health, our
standard of living, our knowledge and the world in which we live. As the
technology progresses, we are finding new ways to use robots. Each new use
brings new hope and possibilities, but also potential dangers and risks. Robotics
is not only a science, but it is also an art. The bots we build reflect the ideas and
personalities we portray. There are many different versions of robots that can be
made. From turtle bots to vehicles like the Mars rovers to rovers like R2-D2.
From walkers that have anywhere from 1 to 10 legs to robotic arms to androids.
Whatever you can dream, you can create. The level of expertise you want your
robot to have and how much learning and research you want to do is up to you.
For those who have relative experience in computer programming and
electronics, this may come easier to you than anyone new to the hobby. Those
who build models, RC vehicles, and other artwork will find it challenging to
modify some of their previous projects.

We have seen how ants always travel in a line, following an invisible route in
search of food, or back home. How on roads we follow lanes to avoid accidents
and traffic jams. Ever thought about a robot which follows line? A perfect or
near perfect mimic of mother-nature? After all the purpose of robotics is to
recreate in terms of machines what we see around to solve a problem or fulfill a
requirement.

Programming intelligence into a robot (or computer) is a difficult task and one
that has not been very successful to date even when supercomputers are used.
This is not to say that robots cannot be programmed to perform very useful,
detailed, and difficult tasks; they are. Some tasks are impossible for humans to
perform quickly and productively. For instance, imagine trying to solder 28
filament wires to a 1/4in square sliver of silicon in 2 s to make an integrated

circuit chip. Its not very likely that a human would be able to accomplish this
task without a machine. But machine task performance, as impressive as it is,
isnt intelligence.

2.2. PROBLEM DEFINITION

In the industry carriers are required to carry products from one manufacturing
plant to another which are usually in different buildings or separate blocks.
Conventionally, carts or trucks were used with human drivers. Unreliability and
inefficiency in this part of the assembly line formed the weakest link. The
project is to automate this sector, using carts to follow a line instead of laying
railway tracks which are both costly and an inconvenience.

2.3. OBJECTIVES OF THE STUDY

The robot must be capable of following a line.
It should be capable of taking various degrees of turns.
It must be prepared of a situation that it runs into a territory which has
no line to follow. (Barren land syndrome)

The robot must also be capable of following a line even if it has breaks.

The robot must be insensitive to environmental factors such as lighting

and noise.

It must allow calibration of the lines darkness threshold.

The robot must be reliable.
Scalability must be a primary concern in the design.
The color of the line must not be a factor as long as it is darker than the
surroundings.

2.4. SCOPE OF STUDY

The robot can be further enhanced to let the user decide whether it is a dark line

on a white background or a white line on a dark background. The robot can also
be programmed to decide what kind of line it is, instead of a user interface. The
motor control could be modified to steer a convectional vehicle, and not require
a differential steering system. The robot could be modified to be a four wheel
drive. Extra sensors could be attached to allow the robot to detect obstacles, and
if possible bypass it and get back to the line. In other words, it must be capable
predicting the line beyond the obstacle. Speed control could also be
incorporated. Position and distance sensing devices could also be built in which
can transmit information to a mother station, which would be useful in tracking
a lost carrier.

2.5. REVIEW OF LITERATURE

First and foremost, no robot could have been built to completion without a
strong hold on the microcontroller used. Most of the basic, intermediate, and
advanced literature about the ATMEGA microcontroller was found in the book
Programming and Customizing the AVR Microcontroller by Dhananjay V.
Gadre. His detailed explanation of every topic made it possible to overcome
many problems which were encountered during design and implementation. The
book also provided a programmer for the ATMEGA microcontroller which was
an indispensable tool helping me experiment with algorithms rather than blindly
copy code from the NET.
The next resource for the ATMEGA microcontroller was the ROBOGYAN by
ROBOTICWARES PVT. LTD. which provides a detailed explanation of each and

every hardware feature and the instruction set. The most helpful is the design
tips section which answered most of the questions which were bogging my head
down.
Robotics by C. K. Kuo was another book I referred for this project. Even though

it was of no direct use to this project, it made me understand many aspects of

robotics which Id have ignored otherwise. Hopefully, the knowledge I have
gained by this book will come in handy in my next robotic endeavor.
Last, but definitely not the least; the Internet. This is where I found websites
giving detailed explanations on a few terminologies. Reference of similar
projects created by others and badly needed tutorials. There is no other place to
easily get the data sheets of the used components. I have listed out a few
websites which were of most help for the project. And in the finale, a huge
thanks to Google.com, no other search engine is even close! No I could not have
done anything without the NET even if I had a million books. The NET is worth
much more.

2.6. APPLICATION
Industrial automated equipment carriers
Entertainment and small household applications.
Automated cars.
Tour guides in museums and other similar applications.
Second wave robotic reconnaissance operations.
2.7. LIMITATIONS
Choice of line is made in the hardware abstraction and cannot be
changed by software.

Calibration is difficult, and it is not easy to set a perfect value.

The steering mechanism is not easily implemented in huge vehicles
and impossible for non-electric vehicles (petrol powered).
Few curves are not made efficiently, and must be avoided.
Lack of a four wheel drive, makes it not suitable for a rough terrain.
Use of IR even though solves a lot of problems pertaining to

interference, makes it hard to debug a faulty sensor.

Lack of speed control makes the robot unstable at times.

2.8. METHODOLOGY
The first idea was to use optical imaging (CCD cameras) to see the line. This
was later given up due to various reasons including complexity and
unavailability of components. Later a choice was made to use an array of
sensors which solved most of the problems pertaining to complexity.
The resistor values used in the sensor array were experimentally determined
rather than theoretical mathematical design calculations. This was done as the
data sheets of the proximity sensor was not available anywhere and most of the
parameters had to be determined experimentally.
The L298 chip is used as it was a much better option than forming an H-Bridge
out of discrete transistors, which would make the design unstable and prone to
risk of damage.
The AVR microcontroller was used as it is the only device I have a full practical
knowledge about, and most of all a RISC processor which are better suited for
real-time operations. Thus the midrange devices were chosen. At PB3 pin pwm
can be generated.
The PCB is the only thing in the whole project for which I turned for outside
help. By the time the design became successful, time had run out. I no longer
had time or resources to create a PCM myself. This was done by JI Electronics.

The project was entirely (other than PCB) designed, created, soldered, tested
and coded by me. For which Im thankful for, as I have learnt much more in the
processes and not to mention the fun had.

3. THEORY
3.1. THE DIFFERENTIAL STEERING SYSTEM
The differential steering system is familiar from ordinary life because it is the
arrangement used in a wheelchair. Two wheels mounted on a single axis are
independently powered and controlled, thus providing both drive and steering.
Additional passive wheels (usually casters) are provided for support. Most of us
have an intuitive grasp of the basic behavior of a differential steering system. If
both drive wheels turn in tandem, the robot moves in a straight line. If one
wheel turns faster than the other, the robot follows a curved path. If the wheels
turn at equal speed, but in opposite directions, the robot pivots.

Figure 3.1: The Differential steering model

Where give the displacement (distance traveled) for the left and right wheels
respectively, is the turn radius for the inner (left) wheel, is the distance
between wheels (from center-to-center along the length of the axle), and is the
angle of the turn in radian. In this discussion, we will treat the axle's center

point as the origin of the simulated robot's frame of reference.

Once we've established the simple geometry for the differential steering system,
it is easy to develop algorithms for controlling the robot's path. Note, though,
that we did make an important simplifying assumption: the wheels maintain a
steady velocity. We neglected the effects of acceleration. If the wheels are
allowed to accelerate, the curve which describes the robot's trajectory can
become much more complicated. When working with very light robots, where
the mass (and inertia) of the platform is small, we can often get away with
treating changes in speed as nearly instantaneous. The path that the robot
follows will not be truly circular, but it will be close enough for many
applications. For larger and heavier robots, of course mass is important and
acceleration must be considered.
If the right wheel is moving at a velocity of VR and the left wheel at a velocity
of VL, then the following equation can be derived :Where a positive implies counter-clockwise rotation; the above equation
clearly shows that the angle of the turn can be increased by either,

Increasing the difference in the wheels velocities (VR VL), or

Keep the wheels at the different velocity for a longer time (t)
All this while b remains constant; in the line following robot, both these
parameters are dynamically changed by the sensors in order to keep the robot on
the line.

3.2. D.C. MOTORS

DC motors are widely used, inexpensive, small and powerful for their size.
Reduction gearboxes are often required to reduce the speed and increase the
torque output of the motor. Unfortunately more sophisticated control algorithms

are required to achieve accurate control over the axial rotation of these motors.
Although recent developments in stepper motor technologies have come a long
way, the benefits offered by smooth control and high levels of acceleration with
DC motors far outweigh any disadvantages.
Several characteristics are important when selecting DC motors and these can
be split into two specific categories. The first category is associated with the
input ratings of the motor and specifies its electrical requirements, like
operating voltage and current. The second category is related to the motor's
output characteristics and specifies the physical limitations of the motor in
terms of speed, torque and power.
Example specifications of the motors used are given below:

As noticed, the torque provided can hardly move 30gm of weight around with
wheel diameter of about 2cm. This is a fairly a huge drawback as the robot
could easily weigh about a kg. This is accomplished by gears which reduce the
speed (2400 rpm is highly impractical) and effectively increase the torque. If the
speed is reduced by using a gear system by a factor of

then the torque is

increased by the same factor. For example, if the speed is reduced from 2400
rpm, to 30 rpm, then the torque is increased by a factor of (2400/30 = 80) in
other words the torque becomes 30 80 2400 gm-cm or 2.4 kg-cm which is more
than sufficient.

3.3. H-BRIDGE MOTOR CONTROL

DC motors are generally bi-directional motors. That is, their direction of
rotation can be changed by just reversing the polarity. But once the motors are
fixed, control becomes tricky. This is done using the H-Bridge. The figure is
given below.

The Explanation is simple, If A & D are turned on, then the current flows in the
direction shown in the figure below.

Figure 3.3: Clockwise rotation

If B & C are turned on, then the motor rotates in counter clockwise direction.

Figure 3.4: Counter-Clockwise rotation

If you turn on the two upper circuits, the motor resists turning, so you
effectively have a breaking mechanism. The same is true if you turn on both of
the lower circuits. This is because the motor is a generator and when it turns it
generates a voltage. If the terminals of the motor are connected (shorted), then
the voltage generated counteracts the motors freedom to turn. It is as if you are
applying a similar but opposite voltage to the one generated by the motor being

turned. In other words, it acts like a brake. Any other state like A & C = ON or
B & D = ON will cause a direct path to ground causing a very high current to
pass through the relays thus causing a burnt fuse (if it exists).
The following figure shows an H-Bridge using only transistors. The same
theory applies.

Figure 3.5: H-Bridge using transistors.

Usually, the above circuitry can be used only for direction control. The Existing
H-Bridge is further modified to include another transistor, now making speed
control possible too. This is shown in the figure below.

Figure 3.6: Enhanced H-Bridge

The same direction rules apply, but now the motor will behave as per the
direction control only when a 1 is given to the EN input. Speed control is
usually done by giving a PWM signal, and the duty cycle is varied to vary the
speed of the motor. Usually protection diodes are also incorporated across the
transistors to catch the back voltage that is generated by the motor's coil when
the power is switched on and off. This fly-back voltage can be many times
higher than the supply voltage! If diodes are not used, the transistors have a
good chance to get burnt.
3.4. THE IR SENSORS

The MOC7811 consists of an infrared emitting diode ( = 950nm) and an NPN

silicon phototransistor mounted to face each other on a converging optical axis
in a black plastic housing. The phototransistor responds to radiation from the
emitting diode only when no object is present within its field of view. This
sensor is physically modified so that the emitter and detector face the same
direction and thus the modified sensor serves the purpose of an opticalreflective sensor. The sensor has a focal length of 8mm, thus the surface must
be at an optimum distance of 1.6cm. The original and modified sensors are
shown below.

Figure 3.7: LEFT: Original sensor, RIGHT: modified sensor

Figure 3.8: Reflective sensor

If a reflective (white) surface is present at the optimal distance (d = 1.6cm) then

the reflected waves will strike the detector which on radiation will start to
conduct. The circuit diagram is shown in the figure below.
3.5. COMPARATOR

A comparator is a circuit which compares a signal voltage applied at one input

of an op-amp with a known reference voltage at the other input, and produces
either a high or a low output voltage, depending on which input is higher. The
input / output characteristics of a comparator is as shown.

Figure 3.9: Comparator transfer characteristics.

3.6. VOLTAGE REGULATOR

It has been shown that practically all electronic devices need DC supply. A
direct voltage of constant magnitude requires to be supplied, for the smooth and
efficient functioning of these devices. A properly designed voltage regulator
ensures that, irrespective of change in supply voltage, load impedance or
temperature, the DC supply is maintained at a constant level. This is achieved
by incorporating some type of feedback in the regulator circuit.
An IC voltage regulator unit contains all the circuitry required in a single IC.
Thus there are no discrete components and the circuitry needed for the reference
source, the comparator and control elements are fabricated on a single chip.
Even the over load and short-circuit protection mechanism is integrated into the
IC. IC voltage regulators are designed to provide either a fixed positive or
negative voltage, or an adjustable voltage which can be set for any value
ranging between two voltage levels.

3.7. BATTERY

Motors on a robot consume most of the power. For most of them, each DC
motor typically consumes 1.5W on the average. For differential steering, two
DC motors consume up to 3W. By comparison, the logic components typically
draw a total of about 80mA. Even at a supply voltage of 12V, the logic
component only consumes 1W.

3.8. INTELLIGENCE
There are two schools of thought concerning the creation of intelligence in
artificial systems. The first approach programs an expert system (top down); the
second is a neural or behavior based system (bottom up). The expert system
uses rules to guide the robot in task performance. Behavior based programs
create an artificial behavior in the robot that causes it to reflectively
(automatically) perform the task required. Behaviors may be

programmed

(software) or may be hardwired into the robot. Behavior based intelligence

doesnt require a central processor, although such a system may have one.
Lets look at a practical programming problem and see how each approach
differs. Suppose you worked for a company that designed a new robotic vacuum
cleaner. The purpose of the robot is to vacuum the floor of a customers home
or apartment. Your job is to program the navigation system. The robot needs to
move autonomously throughout the house. How would you go about
programming the robot to accomplish navigation around the home so it could
travel in and out of rooms without destroying the place?
Lets assume you first decide to try an expert navigation system. This approach
uses brute force programming and a lot of memory. You might begin by
dividing the task of vacuuming the apartment or home into smaller tasks such as
vacuuming individual rooms. You begin by programming into the robots
memory an electronic map (floor plan) of the home or area where the robot
needs to vacuum. Then you map out each individual room and its contents. The
robot must have the ability to measure its movement as it moves as well as
compass direction to maintain its location integrity. Once this is accomplished,
the robot must have an exact start location on the floor plan.
The robots movement from the start position is measured and plotted on its

internal floor plan map. Problems occur if an object is positioned differently or

is out of place, such as a trash receptacle or chair that has been moved. In this
situation the real world does not match the robots internal map. Similar
problems occur if new objects are left on the floor such as a bag, toy, or pet.
Even so, these obstacles would not present too much of a problem for an expert
system. To compensate, a secondary collision detection subprogram could be
written to detect, map, and go around an obstacle not existing on the internal
map. The robot continues to move and vacuum the floor. Keep in mind that as
the robot navigates around new obstacles, its continually updating its internal
map as it travels, to maintain its location integrity. These tasks are gobbling up
computer time and memory.
The robot vacuum accomplished its task. Now suppose you w ant to share this
robot or rent it. Now you have a problem. Each new house and every room in
the new house would require its own electronic map. Although expert
programming does work, it tends to be inflexible and not adaptive toward new
or innovative situations.
Now lets try the other approach that uses behavior based or bottom-up
programming. Instead of programming internal maps, we program sensor
responses and behavior based algorithms (feed-forward and feedback loops) for
sensing and traveling around obstacles and avoiding getting stuck underneath
furniture or trapped in corners. Without any internal map we allow the robot to
travel and move around the house in a random manner. The idea is that while
traveling in a haphazard manner, it will eventually make its way throughout the
rooms, cleaning the floor as it goes. Because the robot travels randomly, it will
take longer for the robot to vacuum the entire floor, and it may miss a spot here
and there, but it gets the job done. Since this behavior based type of robot
vacuum isnt programmed for a particular house or room, it may be used in any

house in any room at any time.

While our example is simple, it does illustrate the main differences between
expert and behavior based (neural) programming. But lets look at just one more
example before we move on.
Expert systems typically have all the answers that the designers believe will be
required by the system programmed into the system before it begins. It may
store and categorize new information, but based on previously determined
categories and existing knowledge. An example of this system could be a rock
identification system. The robot examines unknown rocks based on known
characteristics of rocks, such as color, hardness, scratch ability, acid reaction
tests, mass, etc. The expert system fails if it inadvertently picks up a piece of ice
that melts to water during the tests. Well, it fails as long as the designer(s) never
anticipated the robot picking up a piece of ice by mistake and made allowances
for it.
Neural (behavior based) systems are not programmed and are more adaptive, as
shown in the previous example. But is a neural system suitable for this task of
rock identification? Probably not! There are instances in which expert systems
are the method of choice. One shouldnt blindly assume one system is better
than the other in all cases.
To date, behavior based robots are more successful at task accomplishments
such as traveling over unfamiliar and rough terrain than are programmed robots.
(Other neural based intelligence includes speech recognition, artificial vision,
speech generation, complex analysis of stock market data, and life insurance
policies.)
The line following robot uses behavior based programming, to accomplish the
task at hand. For one, the system may not be a neural system pre say. But this

is simulated in software.

3.9. THE AVR ATMEGA 16 MICROCONTROLLER

4. DESIGN AND IMPLEMENTATION

4.1. SCHEMATIC
The schematic of the Line following robot is shown in the figure. The main
component is the ATMEGA 16 microcontroller. Due to page limitations, the
schematic is divided into two sections; one the Sensor Array Board, and the
other the motor-control or main board.
The main features incorporated into the hardware are given below:

The AVR ATMEGA 16 microcontroller

The voltage regulator [IC 7805]
Electrolytic Capacitor (100uf)
IN 4004 [DIODES]
Crystal oscillator (16MHz)
The H-bridge motor control IC (L298)
Motors (6V), with coupled reduction gears.
4*1.5V, R6 Zinc Chloride battery.
LTH-1550-1 IR interrupt sensor
The LM324 comparator IC
INDUCTOR [10uH]
Connectors to join the different boards to form one functional device.

Each of the hardware is dissected and was designed/implemented separately for

their functional and later incorporated as one whole application. This helped in
the debugging processes. In similar fashion the separate modules forming the
ensemble will be explained separately.

4.1.1. Sensor Circuit Research and Development

The research done on the line sensors showed that there are two main
methods used:
A Light dependant resistor picking up light reflected off the floor from an
LED.
A Phototransistor picking up light reflected from an infra red LED.
The Infra Red option has 2 major benefits over the visual spectrum solution.
Firstly that the Infra Red Spectrum is a lot cleaner than the visual spectrum,
there is a lot less potential for interference. The other advantage is that
phototransistors will react much quicker than Light dependant resistors.
However, what appears white and black to the naked eye, may not be
distinguishable in the Infra red spectrum.
The resistance of the sensor decreases when IR light falls on it. A good sensor
will have near zero resistance in presence of light and a very large resistance in
absence of light. We have used this property of the sensor to form a
potential divider. The potential at point 2 is
Rsensor / (Rsensor + R1).
Again, a good sensor circuit should give maximum change in potential at point
2 for no-light and bright-light conditions. This is especially important if you
plan to use an ADC in place of the comparator.
To get a good voltage swing, the value of R1 must be carefully chosen. If

Rsensor = a when no Light falls on it and Rsensor = b when light falls on it.
The difference in the two potentials is: Vcc * { a/(a+R1) - b/(b+R1) }

Relative voltage swing = Actual Voltage Swing / V

= V * { a/(a+R1) - b/(b+R1) } / V
= a/(a+R1) - b/(b+R1)
cc

cc

cc

4.1.2. THE MICROCONTROLLER

Atmel's AVR microcontrollers have a RISC core running single cycle

instructions and a well-defined I/O structure that limits the need for
external components. Internal oscillators, timers, UART, SPI, pull-up resistors,
pulse width modulation, ADC, analog comparator and watch-dog timers
are some of the features you will find in AVR devices.
AVR instructions are tuned to decrease the size of the program whether the code
is written in C or Assembly. With on-chip in-system programmable Flash and
EEPROM, the AVR is a perfect choice in order to optimize cost and get product
to the market quickly.
-http://www.atmel.com/products/avr/
Apart form this almost all AVRs support In System Programming (ISP) i.e.
you can reprogram it without removing it from the circuit. This comes very
handy when prototyping a design or upgrading a built-up system. Also the
programmer used for ISP is easier to build compared to the parallel programmer
required for many old uCs. Most AVR chips also support Boot Loaders which
take the idea of In System Programming to a new level. Features like I2C bus
interface make adding external devices a cakewalk. While most popular uCs
require at least a few external components like crystal, caps and pull-up
resistors, with AVR the number can be as low as zero!
Cost: AVR = PIC > 8051 (by 8051 I mean the 8051 family)
Availability: AVR = PIC <8051
Speed: AVR > PIC > 8051
Built-in Peripherals: This one is difficult to answer since all uC families
offer comparable features in their different chips. For a just comparison, I would
rather say that for a given price AVR = PIC > 8051.
Tools and Resources: 8051 has been around from many years now,
consequently there are more tools available for working with it. Being a part of
many engineering courses, there is a huge community of people that can help
you out with 8051; same with books and online resources. In spite of being new
the AVR has a neat tool chain (See References and Resources). Availability of
online resources and books is fast increasing.
Here, 8051 > AVR = PIC

4.1.3 CRYSTAL OSCILLATOR

The clock frequency is provided by one 16 Mhz crystal which is connected

across the OSC1 & OSC2 pins as shown above. This provides an instruction
execution time of 1 s.

Fig 4.1
4.1.4. BATTERY

If we assume the whole robot consume 5W, it requires 4500J of energy to last
15minutes. If we use a 6V battery, it must have a capacity of
4500J/6V=750Asec or 104mAH. This may imply that getting a battery of
150mAH is sufficient. Unfortunately, the discharge curve of a 150mAH will not
sustain the required voltage for 15 minutes.
Thus a Zinc Chloride battery was used of rating of 1.2mA for the robot to last
longer than 15min and also to take practical situations into considerations.
4.1.5. VOLTAGE REGULATOR

The circuit requires two voltage sources; one for the digital ICs (+5V) and a
+12V to the motors. The motor is supplied 12V unregulated supply directly
from the battery as regulation would be difficult and unnecessary; whereas the
digital ICs and the microcontroller require a perfect ripple free +5V to function
properly. The L7805C is a 5V voltage regulator IC. The capacitors added to the
input of the voltage regulator are to isolate the spikes generated by the motor

from the input and to reduce noise. The 10 F capacitor at the output is to
maintain stability and improve regulation. These are standard values. The 0.1
F capacitor is used at the input because of the fact that high value capacitors
have poor high frequency response.
4.1.6. D.C. MOTORS

Geared D.C. motors were used which can operate in the range from 0V to 12V
.The D.C. motors have a speed of 2400rpm and a torque of 15gm-cm. The gears
decrease the speed to 30rpm at 6V and thus considerably increasing the torque
so that the robot can carry the load of its frame and the lead-acid battery. Two
such motors are used in the rear of the robot, and a dummy castor is fixed to the
front to stabilize the robot.

4.1.7. THE H-BRIDGE CONTROL HARDWARE

The entire motor control circuitry is shown in the above figure along with the
internal circuitry of the L293D motor control IC. The table below clearly
indicated the operation of the IC.

The total number of directional control signals required is 4; but as it can be

observed in the above table, IN1 & IN2 are complimentary (and so is IN3 &
IN4) that is, both the inputs have to take the opposite states for a safe operation.
This is done by connecting DL to IN1 and

DL to IN2. The same is done to IN3

& IN4. Now we have 1 directional control per motor. The ENABLE of each
motor section is given PWM inputs to further improve on the control. Now,
each motor has a direction control and a speed control. The clamping diodes are
built into the chip which prevent the back EMF generated by the motors to harm
the H-bridge. The inversion was achieved using the 74HCT04 HEX inverter IC

having a slew rate of about 6ns which is negligible compared to the reaction
time of the H-bridge itself.

4.1.8 COMPARATOR
The reference voltage is generated by the 20k POT and given to all the
comparators to the non-inverting input. When the respective sensor is on the
line, the emitted light is absorbed by the line and the transistor is the cut-off
mode, thus a potential of 4.6V is given to the inverting input which is greater
than Vref (which is chosen to be 2.5V), thus the output of the comparator goes
low. When the sensor is not on the line (reflective white surface) the potential
across the detector is usually 0.6V. Thus the output of the comparator goes high
(the non-inverting input has a greater potential). Thus the output of the
comparator goes low only when the sensor is over the line. The comparator is
open collector, and hence a pull-up resistor of 10 k is required at the output.

4.2 Chassis Research and Development

The size of the buggy was determined by the circuitry which is to go inside it.
Once we had got a basic idea of the electronics side of the project we could
then start to construct the chassis. This determined the positioning of the
sensors, motors and the front skid. We decided to position the line follower
sensor in such a way to enable it to handle the tightest corners.
The boards are mounted to the chassis using M4 50mm bolts these bolts hold
the circuits a safe distance from each other. There are 4 lugs in the chassis, and
4 holes in the body, to hold the two together.

4.3. PROCESS EXPLANATION

As shown in the figure above, is a typical situation involved. At every sampled

time the commands executed by the microcontroller is also shown. From the
above figure, it should be clear about the software requirements.

Spiral movement during line find mode.

The surface is sampled every 2.1ms using the timer 0 interrupt routine. This
implies that the line is sampled 476 times in a second. From observations, the
robot travels at a maximum speed of around 10cm/sec. In other words, 47.6
samples are taken per cm, or 4.76 samples per mm. This is more than ever
required! Due to the fact that the robot can remember and follow the previous
task when its sensors do not see a line, enables the robot to trace a sharp turn,
even if in case the robot runs off the line while making the turn.

Robots line approximation

The 8 sensors are connected to PORTA.

You need not connect anything to AVCC and AREF, it is required only if ADC
is used. The L298 Motor Driver has 4 inputs to control the motion of the motors
and two enable inputs which are used for switching the motors on and off. To
control the speed of the motors a PWM waveform with variable duty cycle is
applied to the enable pins. Rapidly switching the voltage between Vs and GND
gives an effective voltage between Vs and GND whose value depends on the
duty cycle of PWM. 100% duty cycle corresponds to voltage equal to Vs, 50 %
corresponds to 0.5Vs and so on. The 1N4004 diodes are used to prevent back
EMF of the motors from disturbing the remaining circuit. Many circuits use
L293D for motor control, I chose L298 as it has current capacity of 2A per
channel @ 45V compared to 0.6 A @ 36 V of a L293D. L293Ds package is not
suitable for attaching a good heat sink; practically you cant use it above 16V
without frying it. L298 on the other hand works happily at 16V without a heat
sink, though it is always better to use one.

4.4 ALGORITHM

4.5 FLOW CHART

4.6 SOURCE CODE

/*****************************************************
Project: Line Follower
Date : 7/12/2009
Author: Pallav Kumar Nayak

Company: BRM IIT

Compiler: CodeVision AVR
Chip type : ATmega16
Program type : Application
Clock frequency : 7.372800 MHz
Memory model : Small
External SRAM size : 0
Data Stack size

: 256

*****************************************************/
//#define debug 1
#include <mega16.h>
#include <delay.h>
#ifdef debug
#include <stdio.h>
#endif
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define
#define

FWD 0xAA
REV 0x55
R 0x22
L 0x88
CW 0x99
CCW 0x66
STOP 0x00
B 0xFF
RSPEED OCR1AL
LSPEED OCR1BL
SPEED0 255
SPEED1 0
SPEED2 0
SPEED3 0
MAX 3
HMAX 1

void move (unsigned char dir, unsigned char delay, unsigned

char power);
unsigned char i, rdev, ldev, ip, delay, dir, power, dirl,
unsigned char history[MAX],hcount=0,rotpow;
#ifdef debug
Unsigned char rep=0,prev=0;
#endif
void main(void)
{

// Input/Output Ports initialization

// Port A initialization

// Func7=In Func6=In
Func1=In Func0=In

Func5=In

Func4=In

Func3=In

Func2=In

// State7=T State6=T State5=T State4=T State3=T State2=T

State1=T State0=T
PORTA=0x00;
DDRA=0x00;
// Port B initialization
// Func7=In Func6=In Func5=In Func4=In Func3=In Func2=In Func1=In
Func0=In
// State7=T State6=T State5=T State4=T State3=T State2=T State1=T
State0=T

PORTB=0x00;
DDRB=0x00;
// Port C initialization
// Func7=In Func6=In Func5=In Func4=In Func3=In Func2=In Func1=In
Func0=In
// State7=T State6=T State5=T State4=T State3=T State2=T State1=T
State0=T

PORTC=0x00;
DDRC=0xFF;
// Port D initialization
// Func7=In Func6=In Func5=Out Func4=Out Func3=In Func2=In
Func1=In Func0=In
// State7=T State6=T State5=0 State4=0 State3=T State2=T
State1=T State0=T
PORTD=0x00;
DDRD=0x30;
// Timer/Counter 0 initialization
// Clock source: System Clock
// Clock value: Timer 0 Stopped
// Mode: Normal top=FFh
// OC0 output: Disconnected
TCCR0=0x00;
TCNT0=0x00;
OCR0=0x00;
// Timer/Counter 1 initialization
// Clock source: System Clock
// Clock value: 921.600 kHz
// Mode: Fast PWM top=00FFh
// OC1A output: Non-Inv.
// OC1B output: Non-Inv.
// Noise Canceler: Off
// Input Capture on Falling Edge
TCCR1A=0xA1;
TCCR1B=0x0A;
TCNT1H=0x00;
TCNT1L=0x00;
ICR1H=0x00;
ICR1L=0x00;
OCR1AH=0x00;
OCR1AL=0xFF;
OCR1BH=0x00;
OCR1BL=0xFF;

// Timer/Counter 2 initialization
// Clock source: System Clock
// Clock value: Timer 2 Stopped
// Mode: Normal top=FFh
// OC2 output: Disconnected
ASSR=0x00;
TCCR2=0x00;
TCNT2=0x00;
OCR2=0x00;
// External Interrupt(s) initialization
// INT0: Off
// INT1: Off
// INT2: Off
MCUCR=0x00;
MCUCSR=0x00;
#ifdef debug
// USART initialization
// Communication Parameters: 8 Data, 1 Stop, No Parity
// USART Receiver: On
// USART Transmitter: On
// USART Mode: Asynchronous
// USART Baud rate: 57600
UCSRA=0x00;
UCSRB=0x18;
UCSRC=0x86;
UBRRH=0x00;
UBRRL=0x07;
#endif
// Timer(s)/Counter(s) Interrupt(s) initialization
TIMSK=0x00;
// Analog Comparator initialization
// Analog Comparator: Off
// Analog Comparator Input Capture by Timer/Counter 1: Off
ACSR=0x80;
SFIOR=0x00;
while (1){
#ifdef debug
if(rep<255)
rep++;
if(prev!=PINA) {
prev=PINA;
printf("%u\r",rep);
for(i=0;i<8;i++)
printf("%u\t",(prev>>i)&0x01);
rep=0;
}

#endif
if(PINA!=255){
rotpow=255;
ldev=rdev=0;
if(PINA.3==0)
rdev=1;
if(PINA.2==0)

rdev=2;
if(PINA.1==0)
rdev=3;
if(PINA.0==0)
rdev=4;
if(PINA.4==0)

ldev=1;
if(PINA.5==0)

ldev=2;
if(PINA.6==0)

ldev=3;
if(PINA.7==0)

ldev=4;
if(rdev>ldev)

move(R,0,195+12*rdev);
if(rdev<ldev)
move(L,0,195+12*ldev);

if(rdev==ldev)
move(FWD,0,200);
}
else
{
for(i=0,dirl=0;i<MAX;i++) {
if(history[i]==L)
{dirl++;}
}
if(rotpow<160) {rotpow=160;}
if(rotpow<255) {rotpow++;}
if(dirl>HMAX)
{move(CW,0,rotpow);}
else
{move(CCW,0,rotpow);}
}
};
}

void move (unsigned char dir, unsigned char delay, unsigned

char power) {
PORTC=dir;
if(dir==L || dir==R) {
hcount=(hcount+1)%MAX;
history[hcount]=dir;
}
LSPEED=RSPEED=255;//power;
//delay_ms(delay);

5. RESULT AND CONCLUSION

The Line following robot was finally completed. A lot of effort was put into the
design, implementation and days of toil in front of the computer, writing and
debugging the code. The robot was finally running with a few glitches here and
there which were sorted in the later revisions of the firmware. The line
following robot still has a few short-comings but achieves most of the
objectives.
I earned a lot of knowledge on micro-controllers, a deeper & clearer view of the
architecture, ports & all other functional blocks was achieved. Did a lot of
research on robotics and already have my next project planned. Had a peek look
at all simple functional parts of the project like the crystal oscillator, logic gates
and the works. Well, these were the topics that we have already dealt with, but I
must be honest and admit that there were various practical issues which one
would learn only during a project. Theres a lot of learning & yet not the end,
learning is a continuous never ending process but is definitely fun.