This article was downloaded by: [The Aga Khan University]On: 21 November 2014, At: 02:14Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Intelligent Automation & Soft ComputingPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tasj20

Designing and Developing Simple Low Cost MobileRobots for Autonomous NavigationL. Capozzo a , G. Attolico a & G. Cicirelli aa Istituto di Studi sui Sistemi Intelligenti per l’Automazione – C.N.R. , Via G. Amendola ,122/D-O - 70126 , Bari (Italy) E-mail:Published online: 01 Mar 2013.

To cite this article: L. Capozzo , G. Attolico & G. Cicirelli (2006) Designing and Developing Simple Low Cost Mobile Robots forAutonomous Navigation, Intelligent Automation & Soft Computing, 12:2, 173-181, DOI: 10.1080/10798587.2006.10642923

To link to this article: http://dx.doi.org/10.1080/10798587.2006.10642923

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Intelligent Automation and Soft Computing, Vol. 12, No. 2, pp. 173-181, 2006 Copyright © 2006, TSI® Press

Printed in the USA. All rights reserved

173

DESIGNING AND DEVELOPING SIMPLE LOW COST MOBILE ROBOTS

FOR AUTONOMOUS NAVIGATION

L. CAPOZZO, G. ATTOLICO, G. CICIRELLI Istituto di Studi sui Sistemi Intelligenti per l’Automazione - C.N.R.

Via G. Amendola, 122/D-O - 70126 Bari (Italy) E-mail: capozzo@ba.issia.cnr.it

ABSTRACT—The aim of this article is to illustrate how it is possible to design and build low cost mobile robots by using only simple commercial electronic devices and toys. As suggested by Valentino Braitenberg [1] in his experiments on synthetic psychology we developed two simple and compact mobile robots, each capable of doing two different autonomous behaviors: the wall following (moving the robot at a constant distance from the wall on its left or right hand) and the obstacle avoidance (avoiding collision with obstacles during the path). Two different control systems have been developed and compared. The first system is controlled by an analogical circuit obtained by using electronic components (transistors, diodes, resistors, capacitors …) whereas, the second is controlled by a digital circuit which employs a micro-processor board. Those platforms allow the robot configuration to be easily adapted to the specific requirement of each application. Their strengths and limitations are also detailed. Both the approaches are useful to reach the main purpose of our work which is the development of platforms to help students to learn electronics, computer programming and robot control [5]. Following this approach learning becomes a practical and, most of all, playful activity. Key Words: Mobile robots; Analogical control; Digital control; Wall-following behavior; Obstacle-avoidance behavior

1. INTRODUCTION The increasing interest in micro-controller boards together with the drastic reduction in the cost and

the size of integrated circuits have opened a new field in the development of autonomous systems. This new trend in design and building autonomous vehicles is considered by most researchers an art as much as a science. The large diffusion of autonomous systems in many activities has inspired many researchers all around the world. The human security in dangerous works, support to disabled and repetitive actions, are some of the fields where robots can be employed.

In this article we illustrate how it is possible to construct low cost vehicles capable of doing some basic behaviors using commercial toys and electronic components. Our work follows in spirit a research activity extensively studied by V. Braitenberg [1], [2]. In his works he describes how it is possible, starting from the interaction of simple motors and sensors, to create autonomous agents capable of doing particular behaviors. He illustrates a set of experiments in which increasingly complex vehicles are built from simple mechanical and electronic components. Each of these vehicles reproduces an intelligent behavior. The use of the term "intelligent" is strictly related to the way by which the robot performs a particular behavior. We are referring to reactive intelligence as a proper action is associated to every sensory situation.

LEGO1 bricks have been used to develop the structure of the vehicles. With these plastic elements it is possible to build flexible and robust structures for the robots; besides, the particular shape of the single bricks allows the easy integration of sensors and motors as well. The shape of the bricks allows also the optimization of the physical configuration of the mobile robot (number, type and arrangement of motors and sensors) in order to meet the requirements of each specific application. Using both LEGO bricks and simple electronic devices such as relays, resistors, diodes and capacitors, two different robots have been

1 Trademark or registered trademark of LEGO Systems, Incorporated.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

174 Intelligent Automation and Soft Computing

built and compared using two testing behaviors: the wall following and the obstacle avoidance. The robots have two different control systems: the first is obtained using only analogical components like resistors, diodes, transistors and so on; the second is completely digital and controlled by a microprocessor board. The two considered behaviors have been implemented on both vehicles. A touch sensor for the bumper and some infrared receivers and transmitters diodes have been used. The choice of infrared sensors has been motivated by considering both their dimension and their simple electronic interface. A simple controller circuit has been designed to control the analogical vehicle; a controller board, developed at the Massachusetts Institute of Technology, has been programmed for the digital vehicle.

Comparing the two robots we point out that the analogical robot may be easily designed for simple behaviors while it can be unsuitable for complex control architecture. It exhibits the advantages of continuous systems but is sensible to the variations of light and environmental interferences. The digital robot, instead, is more robust and it can also be scaled more easily to implement more sophisticated behaviors and control architectures. The rest of the paper is organized as follows: the first section describes the structure of the two vehicles, in particular the LEGO bricks used and the traction system implemented; the second section describe the analogical vehicle developed and the electronic circuit tested; the third section describes the digital vehicle and the program developed; finally the experimental results and the conclusion are illustrated.

2. LEGO PLASTIC BRICKS To build the structure of the two vehicles, we have used LEGO plastic components. The components

of LEGO Technics are fun to play with and can be flexibly and easily combined to construct several different objects. The presence of gears allows the integration of the movement directly into the structure. The size of the LEGO bricks can be expressed in millimeters. In particular the smallest brick is about 8mm long. The LEGO gears can be used to move the structure. An interesting thing to notice about LEGO gears is the diameter which indicates the space requested to mesh them together. The natural unit for the size of LEGO gears is the horizontal LEGO spacing unit. Nowadays the gear size is expressed in gear teeth (8 - 16 - 24 - 40). Gears control the rate at which the mechanical energy is transmitted and converted into motion. In particular the high speed and low torque of the motors are converted into the low speed and high torque necessary to move the vehicle. In our experiments different gear configurations have been tested to identify the torque and speed that match at the best the application needs. One very important strength of the LEGO tools is the facility with which their configuration can be changed: motors, gears and sensors can be easily changed in number, type and arrangement. This enables not only the experimental comparison of different physical configurations but also a tight relationship between the robot and the application at hand. This is a very important point since the development of intelligence, as long as its effectiveness and efficiency, are strongly related to the physical structure of the robot.

For the analogical vehicle we have chosen the 8-24 configuration while for the digital one the (8-24, 16-16, 8-8-8) configurations have been implemented. The use of more gears assures more stability to the axis of the wheel. A differential guidance system has been realized to produce movements. The differential scheme consists of two wheels on a common axis, each driven independently. This configuration allows the robot to move straight, to turn in place, and to follow an arc. One of the major problems with a differentially driven robot with two wheels is its balance. To ensure the right attitude some additional supports must be provided. Usually this is done by mounting on the robot one or two additional free wheels or an external support. Figure 1 shows the solutions that we have designed for our two robots: the analogical version uses a third free wheel while the digital robot uses an external support. Another problem with differential drive is the difficulty to make the robot move straight. In fact, even when the same voltage is applied to the two motors, they turn at slightly different speeds. To deal with this matter the analogical vehicle uses a variable resistor to dynamically regulate the current flowing to the motor. To achieve the same result, the digital vehicle uses a specifically implemented software routine.

3. ANALOGICAL GUIDED SYSTEM To build an analogically controlled vehicle means to design and to develop a system without any

software contribution. This vehicle is able to realize a meaningful behavior only from the interaction of simple analogical devices.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

Designing and Developing Simple Low Cost Mobile Robots for Autonomous Navigation 175

Figure 1. The two developed vehicles: the robot on the left is equipped with a board with the analogical control system; the vehicle on the right is controlled by a digital programmable board.

The electronic hardware of this vehicle has been developed in our laboratory. It is composed of the following items: - A multi point breadboard that includes all the designed circuits; - Micro mechanical switches used to create touch sensitive bumpers; - Infrared (IR) detectors composed of an emitter and a receiver of IR. The emitters send impulses of

infrared light in a range of 0-30cm. These impulses are captured from the receivers and permit the activation of a group of relays that control the current flowing to the motors;

- 9V LEGO motors. Varying the current flow it is possible to control the speed of the motors; - The whole system is powered by two 9V batteries, one for the motors and the other for the electronic

components; - The structure is build up with standard LEGO Technics pieces. This means that the whole construction

(including the chassis, gears, wheels and the plastic bumper) is made by LEGO. Two micro touch sensors are used to realize the bumper (see figure 2a). With these components it is

possible to detect accidental crashes during the movement. Two mechanical micro-switches are connected in parallel using a plastic pole. They connect the power system to a temporized circuit that controls the direction of movement of the vehicle. The temporizing circuit is a RC cell. By choosing the value of the capacitor it is possible to control its unloading time. When a pressure is applied on the bumper, one of the micro-switch closes the circuit and the current runs through the RC cell. The current loads the capacitor which starts to unload on the resistor. At the same time the relay switch closes and the motors goes backwards. At the end of the unloading time the current flow becomes null and the relays switches to their previous connection. The sensitivity of the bumper is determined by the deflection of the pole keeping the bumper in contact with the micro mechanical switches. In particular three different directions of

Figure 2. a) Bumper circuit; b) IR sensors circuit.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

176 Intelligent Automation and Soft Computing

collision can be detected: left, right and frontal. A couple of IR sensors has been mounted in front of the vehicle as a proximity sensor. They are able to detect an accidental obstacle before the activation of the bumper and are used for a double purpose. The first is to prevent collisions, the second to detect the attitude of the vehicle with respect to the frontal wall during wall following. Without these sensors the vehicle would be unable to detect a frontal obstacle until its contact with the bumper.

The IR sensors allow the vehicle to estimate the distance from the wall on the left hand and to control the robot heading for keeping the vehicle parallel to it. This kind of reaction is obtained by analogically processing the IR signals provided by the detectors. When a receiver captures a signal the corresponding relay switches on and the current flow runs through a resistor connected to a motor which decreases its velocity: therefore the vehicle curves accordingly. The wall following is obtained using two sensors, each composed by a transmitter and two receivers. Figure 3 shows the strategy used for the wall following behavior. Dotted lines indicate the two minimum thresholds from the wall. The vehicle must follow the wall keeping its distance from it between the two thresholds (as illustrated in position 5).

Figure 3. The wall following strategy: the vehicle must keep its distance from the wall between the dotted lines.

In position 1 the couple of front IR sensors situated on the left side receives the reflection from the wall and a little turn on the right is performed. When also the back left IR sensors reach the same threshold another little correction on the right motor occurs. If the frontal IR sensors exceed the second threshold another correction occurs until the vehicle is in position 5.

The use of two different IR receivers is motivated by the two thresholds. Two different resistors permit to fix the value of the thresholds. By using the circuit shown in figure 2b) it is possible to understand the behavior of the electronic circuit. Normally motors receive current directly from the 9V supply battery. When the first receiver captures the IR signal the current flows through the transistor and the relay switches on. In this condition the current flows through the resistor and as a result the motor decrease its velocity. If also the second receiver captures the IR signal the current flows through the second transistor and the second relay switches on. The current to the motor flows through two resistors and as a result the velocity of the motor further decreases. The polarization current is obtained by varying the resistor on the receiver. Using the same circuit the obstacle detection is easily obtained. When the frontal receiver captures the IR signal the relay switches on and the motor inverts its speed.

4. DIGITAL GUIDED SYSTEM The electronic circuit that controls the analogically guided vehicle served its purpose quite well. Using

only relays, potentiometers, mechanical switches and some other discrete components the analogical vehicle is able to avoid obstacles and to follow walls. This approach has been compared with the more flexible system, based on a microprocessor board, used for the digital vehicle. The different components of the digital system are all implemented by suitably routines written in the programming language available for the board.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

Designing and Developing Simple Low Cost Mobile Robots for Autonomous Navigation 177

Changing behaviors in the analogically guided vehicle requires the adjustment of potentiometers, the re-wiring of the circuit, the addition or modification of the components. Changing behaviors in a digital guided vehicle, instead, needs a modification of the software.

The main components of the used controller board are: the processor that executes instructions, a memory in which the results of these executions are stored, ports which interface the computer to different devices (sensors, motor controllers, …) and a bus which provides the communication channel between all these elements. A very important feature of the controller board is the control of sensors (input) and of actuators (output). The board offers digital and analogical sensorial inputs and can control up to six DC motors simultaneously. The operating system executes programs written in IC (Interactive C) which are loaded by the processor using a simple command. The board has been developed at the Massachusetts Institute of Technology: it is distributed in kit and needs to be assembled. The structure of the vehicle is quite similar to the analogical version: an important difference is the substitution of the third free wheel with an iron ball that reduces the friction of the floor. The use of a microprocessor board involves the development of a specific software to produce the desired behaviors. Generally this kind of board requires both some programming in assembly language and some higher-level language routines. The assembly language programming consists of writing code in the machine-specific instruction set designed for the microprocessor. This code is a set of mnemonic instructions, called assembler, which creates the bit level sequence that can be down-loaded to the microprocessor. To create the vehicle behaviors, a high-level language such as C is often used. In this work we have used the IC (Interactive C) that was developed by Randy Sargent and Fred Martin of the MIT Media Laboratory [4]. This software runs on the 6.270 micro-controller board and includes several useful features such as the ability to initiate and terminate processes and to execute C statements immediately, without the need of compiling, linking and loading. Using this software a program has been developed to produce the same behavior of the analogical vehicle. The software strategies used to write the C code for the digital robot is based on its multitasking capability. The software is composed by a series of processes that runs simultaneously with different priority. The board supports this architecture with functions that are able to create a process, to define its time slice and to kill or suspend it. These functions are: Start_process(), Defer() and Kill_process(). Start_process creates a process with a default slice time of 5ms. This function gives in output an identifier that can be used to manage the process itself. Defer suspends temporarily a process for a time slice and restarts it after the defined time. Kill_process kills a created process. The logical behavior of the vehicle is organized on two main levels: the first one realizes the Obstacle_avoidance behavior using the touch sensors and the infrared sensors situated on the front of the vehicle; the second level is the Wall_following behavior and uses the lateral infrared sensors. The management of processes has been developed using a priority based architecture. Behaviors arises from the coordinated interaction of three concurrent processes called: Crash(), Obstacle_detect() and Wall_detect().

The Crash process has the maximum priority: it has the task of detecting an accidental crash during the path. In this case the vehicle is able to go away from the obstacle using a time slice in which no external processes can interfere with the control. The Obstacle_detect has a medium priority. It has only the possibility of interrupting the Wall_detect process. The Wall_detect process, instead, has the minimum priority level and it can be interrupted at any moment (see figure 4). All these are called input processes. They read signals from the sensors, process these signals and set accordingly a variable called Action. A fourth process, called Motor_drive(), read the value of the Action variable and run the correspondent

Figure 4. The priority based strategy

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

178 Intelligent Automation and Soft Computing

command. Motor_drive is an output process that drives the motors. The whole system is controlled by a scheduling scheme which assign to each process a defined time slice for the execution before switching to the next process. The priority management is realized by setting a global variable Level that any process reads at the beginning of every time slice and written in output.

Processes communicate using these global variables Action and Level. Each possible value of the Action variable defines a specific velocity pair for the two motors. This variable is written by the processes Crash, Obstacle_detect and Wall_detect and read by Motor_drive. The Level variable defines the priority level. Each process checks continuously if the current value of the variable Level is not higher that its own priority to decide if it can start its execution. The Motor_drive is always executed and before exiting it sets the priority Level variable to the minimum value before restarting the cycle.

The main() function runs all processes with a user defined time slice until a Kill_process command occurs. For example, after checking the priority level the Crash process reads the digital port on which the two micro-switches of the bumper are connected (see figure 5): if its level is high, Action and Level are set and the task is interrupted. The same is done by the Obstacle_detect function that reads the port associated to the frontal IR sensor. The Wall_detect process, instead, sets the Action variable using a strategy that depends on the values read from the infrared sensors on the left side. Wall_detect process is allowed to set the Action variable if the Level priority is correct. The value of this variable depend on the combination of the sensorial signals produced by the lateral sensors, returned by the function Double_ir_read(). Each sensor is able to assume two state: 0=Not detected; 1=Detected. The combinations of each couple of sensors can be : 00, 01, 10, 11. Considering the value of the signals we should have the following values of the variable Ir1 (or Ir2): 1 = No detection; 2 = Detection of one of the two sensors; 3 = Detection of both sensors (as illustrated in table I). The possibility related to the detection of only the farthest sensor (S1=0 and S2=1) is impossible considering the real task. In this condition the value of the variable Ir1 should always be 2 and the value detected by the function Double_ir_read would be valid. The Action variable has 9 conditions (see table II). For example considering the vehicle near to the wall both the frontal sensors are in the threshold line. The output of the sensorial couple is: S1=1 and S2=1. For the back sensors, instead, we have: S1=1 and S2=0. Considering:

Figure 5. IC code of the Crash process

Ir =S1+S2+1 we have:

Ir1 =1+1+1=3 and

Ir2 =1+0+1=2

Table I. Sensors combination used to implement the wall-following behavior. The error condition means an impossible spatial configuration: the farthest sensor S2 cannot detect the wall if the nearest one S1 does

not detect anything.

S1 S2 Ir=(S1+S2)+1 Error 0 0 1 1

0 1 0 1

1 2 2 3

No Yes No No

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

Designing and Developing Simple Low Cost Mobile Robots for Autonomous Navigation 179

Table II. Action Table: it relates the sensor readings to the values of the variable Action and to the command given to the motors of the vehicle.

Ir1 Ir2 Action=(Ir1+Ir2)*Ir2 Command 1 1 1 2 2 2 3 3 3

1 2 3 1 2 3 1 2 3

2 6

12 3 8

15 4

10 18

Straight Left Left

Right Straight

Left Right Right Right

The action variable is:

Action =(Ir1+Ir2)*Ir2=(3+2)*2=10

This value is able to set the velocity of the motors to cause a little turn on the right. The expression of the variable Action is defined in order to have only a value for each combination as illustrated in table II. The main program is automatically loaded by the board using a download routine of the interactive C. By pressing the red button of the board the vehicle is able to perform the task until the escape button is pressed.

5. EXPERIMENTAL RESULTS In order to test the vehicles a selected path have been performed by both the guidance systems. The

main purpose of this test is not to confirm the best performance of the digital vehicle but to illustrate how it is easy to implement the same development concept in two simple ways. The main differences between the two robots are both mechanical and electronic.

From a mechanical point of view the vehicles differ in the equipment used for obtaining their balance and in the gears system used to transmit the energy from motors to wheels. In particular the analogical vehicle uses a third free wheel for balance and two gears, one connected to the motor and the other to the wheel: the ratio 8-24 changes the high speed low torque movement of the motor to the low speed high torque of the wheel. On the digital vehicle an iron ball, which exhibits a lower friction than the several rolling frictions associated to the third free wheel, has been mounted for balance. Moreover the iron ball has a significantly lower interference with the control of the heading.

To develop the sensory system for the analogical and digital robots some tests have been done in our laboratory under controlled environmental light. Regarding the analogical vehicle, after comparing 5 couples of sensors, we have chosen the couple diode and photo-transistor with linear variation of the output voltage with respect to the transmitted light. These sensors are influenced by two major factors: the reflectivity of the obstacles (which depends on color and kind of surfaces) and the ambient light. For this reason we have chosen white walls and obstacles as to have the best response of sensors. A simple controller circuit has been designed to control the input signal of the photo-transistor and the output to the motors. The sensorial system of the digital vehicle, instead, is more robust than the analogical one. Using the digital trigger of the board a modulated signal has been used as to transmit IR pulses. The result of this strategy is that the receiver captures only the modulated pulses and for this reason is more robust with respect to environmental interferences. The analogical system is more sensible to ambient light and the power dissipation is higher than the digital one. Some tests have been done using different wall shaping (concave wall, convex wall, straight wall) and different environments (with and without obstacles). The experiments have pointed out that the digital vehicle succeeds the 90% of time in doing both the behaviors, whereas the analogical vehicle has a worse performance, in fact it works well only for the 50% of time.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

180 Intelligent Automation and Soft Computing

Figure 6 shows the environment used during the experimental tests.

6. CONCLUSIONS In this article two platforms for helping students to

learn electronics, computer programming and robot control have been presented. The LEGO bricks have been used for building the physical body of the vehicles. They represent an easy and flexible tool for adapting the physical shape and the arrangement of motors and sensors to the requirements of each specific application. This is a very useful feature to investigate the relationship between the body of an agent and the development of its intelligence. Two different architectures have been used and compared for controlling the vehicle: one uses only electronic components and the other relays on a programmable board on which several software modules interact to implement the control behaviors. Both of them provide a practical and playful tool to learn electronics and computer programming even if the digital platform is more suited to experiment with more complex and high levels behaviors.

The main problems encountered with LEGO components are related to the battery capacity and the

motors. Using LEGO motors we have noticed a higher power dissipation and a smaller deflecting torque than in other motors. In order to improve the system we think necessary a reduction of the robot weight and an increase of the sensory system capacity. Future investigation on this subject are related to the improvement of the digital robot in order to develop a cooperative behavior between two or three agents.

Figure 6. The wall following setup

REFERENCES 1. V. Braitenberg. Vehicles: Experiments in Synthetic Psychology. The MIT Press, Cambridge,

Massachusetts, 1984. 2. D. W. Hogg, F. Martin, and M. Resnick. Braitenberg creatures. Technical memo, MIT Media

Laboratory, Cambridge, Massachusetts, 1991. 3. H. H. Lund, O. Miglino, L. Pagliarini, A. Billard, and A. Ijspeert. Evolutionary robotics - a

children's game. In Proc. of IEEE 5th Int. Conference on Evolutionary computation, 1998. 4. F. Martin. The 6.270 robot builder's guide. Technical reference, M.I.T. LEGO design

Competition, Massachusetts. Institute of Technology, 1992. 5. M. Resnick, F. Martin, R. Sargent, and B. Silverman. Programmable bricks: Toys to think with.

IBM Systems journal, (35): 443- 452, 1996. 6. J. S. Albus. The engineering of mind. Intelligent systems division, National Institute of Standards

and Technology, 1996. 7. J. L. Jones and A. Flynn. Mobile Robots: Inspiration to Implementation. A.K. Peters, Wellesley,

Massachusetts, 1993. 8. G. McComb. Robot builder's Bonanza. McGraw Hill, Philadelphia, 1987. 9. M. Resnick. Turtles, Termites, and Traffic Jams - Explorations in massively parallel micro-

worlds. A Bredford Book, The MIT Press, Cambridge, Massachusetts, 1994.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

Designing and Developing Simple Low Cost Mobile Robots for Autonomous Navigation 181

ABOUT THE AUTHORS L. Capozzo obtained the degree in Administration Science from the University of Bari (Italy). From 1990 to 1993 he has been a technician at the Institute for Signal and Image Processing in Bari, under grant of the Italian National Research Council (C.N.R.). Since 1998 he is a staff member at the Institute of Intelligent Systems for Automation of the C.N.R. in Bari. His current research interests cover robotics, electronics devices and the development of simple robots.

G. Attolico received the degree in Computer Science at University of BARI (Italy) in 1986. He is currently senior researcher at the Institute on Intelligent Systems for Automation of the National Research Council. His main interests are in Signal and Image Processing and Analysis, Computer Vision, 3D Modeling and Haptic Interfaces, with application to robotics, visual inspection, image retrieval and advanced interfaces. He is a member of the IAPR and AI*IA.

G. Cicirelli obtained the degree in Computer Science from the University of Bari (Italy) in 1994. From 1995 to 2001 she has been a young researcher at the Institue for Signal and Image Processing in Bari, under grant of the Italian National Research Council (C.N.R.). She is currently a technologist researcher at the Institute of Intelligent Systems for Automation of the C.N.R. in Bari. Her current research interests cover vision-based robotics, pattern recognition, neural networks and reinforcement learning.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

02:

14 2

1 N

ovem

ber

2014