International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 2, July August 2012 ISSN 2278-6856

Autonomous Mobile Robots

S.Benaissa1, H.Medromi2 and S.Tallal3
National High School of Electricity and Mechanics, Hassan II University Team Architecture of Systems, ENSEM, BP 8118, Oasis, Casablanca, Morocco
1, 2, 3

Abstract: Agents are systems acting on some environment,

including their minds. In this case, their action consists of taking decisions, manipulating information, etc. More generally, they act on the physical or social environment producing transitions from one state to another. In this contribution, agents are essentially autonomous, operate without the direct intervention of humans or others, and have some kind of control over their actions and internal state. Agents can be placed on a continuous dimension of autonomy, as several intermediate cases occur between full autonomy and full slavery. Indeed, limited autonomy is the paradigmatic case of humans. Sources of limited autonomy can be found both in resource scarcity and in agents sharing a common environment. Due to resource scarcity agents may not possess all of the means needed for achieving their goals. Robots, unlike many software multi-agents systems, operate under real-world, real-time constraints where sensors and effectors with specific physical characteristics need to be controlled. To facilitate research in autonomous robotics and help architecture designers in managing the complexity of embodied agents, in this paper an autonomous robot control architecture have been developed that support various aspects of the agent development process, ranging from the design of an agent architecture, to its implementation on robot hardware, to executing it on the robot.

management of sensors, an agent for handling the drive and a last agent to act as coordinator of the multi-agent system. Summarizing the SMA is composed of the following agents: Node Agent (NA), a set of sensors. Task Agent (TA), a set of servo motors. Agent System (AS), coordinator of the system.

Keywords: Robotics, Autonomous Mobile Robots, MultiAgents System.

Figure 1 Architecture of the Khep-III complete system. The Node Agent (NA) will be responsible for the sensors used by our robot that are: Ultrasonic Sensor Light Sensor left Light Sensor right Camera These sensors are used to interact with the world to achieve our goal. The Task Agent (TA) will be used to drive to move forward, rewind, or make some movement like escaping. The System Agent (SA) will be responsible for coordinating the other actors, the AN and AT, and will also have activities not only the coordination, which will be to recognize the object that is opposite, calculate the angle of the object, which will help us to know that will guide the robot [2]. Depending on the recognition that was obtained, whether or not the object, which will tell us that there is no escape or the object based on this decision was to trigger the switch which will control the outcome to be used, reactive or path to better performance of the controllers.

1. INTRODUCTION
Globally, any robot is basically designed to replace or to assist the human being in the achievement of a great number of tasks. The complexity of those tasks requires from the robot to have the more and more autonomy at both the decisional and the physical levels [1]. This increasingly requested autonomy is particularly necessary when the robot must evolve/move in an unknown environment, no matter it is structured or not. We highlighted the various aspects that this autonomy aspect should take and thus we precised the functions to include in order to ensure autonomy to a robot. We have been thus led to be working on the implementation of two types of architecture: Structural architecture and the operational architecture of the robot. During the development of our software, we proposed its kinematic characteristics.

2. CONTROL ARCHITECTURE
To realize this work, the creation of a SMA is needed, which is composed of three agents, an agent for the Volume 1, Issue 2 July-August 2012

Page 22

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 2, July August 2012 ISSN 2278-6856
In Figure 1 we show graphically the architecture of the complete system. We can describe the proposed method as follows: In the operation of the model, we have 2 main blocks, which are responsible for knowledge and learning (paradigm of intelligent agents), and the vision and control (fuzzy logic). These modulus are described below, the first module contains 3 agents, NA [Administrative Agent], TA [Task Agent] SA [System Agent] Agent System (AS), and will know every time the operation of the other agents. To detect a change in the environment the Agent Node [that is in charge of the sensors (ultrasonic, camera and two light sensors)] with the ultrasonic sensor and two light sensors, starts its operation, and we start at the state called reactive control; this because you have to move and sense all the time until it finds an obstacle, this can happen once a photo is taken, which will be sent to the database of images (BDI), the image will be applied a pre-processing for recognition in order to know whether the object is found, this decision was taken on the angle (angle to take the decision to bypass the object), able to escape and move forward, trend data, they are caught in the trajectory control process, to observe the speed values [3]. The Task Agent (which is responsible for the drive), once all the necessary parameters for the performance of the robot agent system (AS) who is the coordinator, and executes instructions in the robot. For the development of the Multi-Agent System we used Gaia and UML for the analysis and design of the agents. The completion of the Multi-Agent System (MAS) is based on the FIPA standards for better utilization and greater possibility of extension. 2.1. Agent structure Let's first deal with the notion of intelligent agents. These are generally defined as "software entities", which assist their users and act on their behalf. Agents make your life easier, save you time, and simplify the growing complexity of the world, acting like a personal secretary, assistant, or personal advisor, who learns what you like and can anticipate what you want or need. The principle of such intelligence is practically the same of human intelligence. Through a relation of collaborationinteraction with its user, the agent is able to learn from himself, from the external world and even from other agents, and consequently act autonomously from the user, adapt itself to the multiplicity of experiences and change its behavior according to them. The possibilities offered for humans, in a world whose complexity is growing exponentially, are enormous [42][43][44][45]. The first activity was to perform the analysis and design of the multi-agent system as mentioned above, and to Volume 1, Issue 2 July-August 2012 develop the tool we describe the following concepts of our agent in C++ language.
Agent
#include tache1.h #include tache2.h #include tache3.h #include tache4.h Main ( ) If (condition C 1 ) { Task 1(); } If (condition C 2 { Task2(); } If (condition C 3) { Task3(); } If (condition C 4 ) { Task4(); } . . }

Condition C1 Read / Write

Task1.h
lecture/criture

data1.txt
Condition C2 Read / Write

Task2.h
lecture/criture

data2.txt
Condition C3 Read / Write

Knowledge Base (.h)

Task3.h
lecture/criture

data3.txt Task4.h
Condition C4 Read / Write
lecture/criture

data4.txt

Figure 2 Architecture of System Agent. 2.2. Analysis and Design of the Multi-Agent System (MAS) The first activity was to perform the analysis and design of the multi-agent system as mentioned above, and to develop the two tools we describe the following concepts. 2.2.1. Use Case Diagrams The Use Case diagrams show the granularity of the system into reusable pieces of functionality, interaction of players with the functionality of the system, visually organize user requirements and allow the contract to certify the functionality, and formalize the process map. Figure 3 presents the use case of the sense in which the distance is appreciated that other processes must be carried out before and after to complete.

Figure 3 Use case diagram of the process of image recognition. In Figure 4 shows the use case of the motor A move, which is seen in all other processes that must be carried out before and after to complete [4]. Page 23

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 2, July August 2012 ISSN 2278-6856

If the distance is < di

Stop Figure 4 Main processes of use cases.

Figure 5 presents the case for use of the remote sensing process, which shows that other processes must be carried out before and after to complete.

1.

GUltrasonic

Move Motor A,(TA)

Figure 7 Sequence diagram use case to move motor A. In Figure 8 presents the sequence diagram of use case to capture the image which is carried out by the agent node (AN). Take Picture (NA) 3. DeteUltrasonic
Off

Figure 5 Use case diagram of the remote sense.

On

Figure 6 shows the use case of the motor A move, which is seen in all other processes that must be carried out before and after to complete.

Take Picture

Figure 8 Sequence diagram use case capture image. In Figure 9 we show the sequence diagram used for the case sense of distance which is carried out by the agent node (AN) [5], [6]. Figure 6 Use case diagram of the process Moving motor A. Sequence diagrams describe the interaction of objects that require the functionality of the different scenarios of a use case, objects are represented with their life cycle within a time series, and each possible scenario of a use case can be represented as a sequence diagram. Below in Figure 7 we show the sequence diagram of the use case for moving motor A.

Volume 1, Issue 2 July-August 2012

Page 24

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 2, July August 2012 ISSN 2278-6856
Permission: Depends entirely on the agent system to perform its task. Activities: Sensors for the engine: they measure in degrees is given for each engine is 360 degrees for one revolution, a revolution for the tire of each motor, the motors can rotate independently, can also be used with constant speed motors.
- Ultrasonic - Infrared

Figure 9 Sequence diagram use case sense distance. 2.3. Implementation of our architecture of mobile robot Below we show the diagrams in which we can observe the responsibilities, permissions, activities and protocols to be followed by each agent in the Multi-Agent System (MAS). Agent node Responsibilities: Sense: Active Inactive Permissions: Ultrasonic sensor: measures distances from 0 cm to 255 cm, with a delay of one millisecond signal. Camera Sensor: take pictures in an estimated time of seconds, saves all images in a folder for processing. It has a degree of vision approximately of 50 degrees. Light sensors: they measure the intensity with which an object reflects light. This makes a light emitting and measuring the portion of the return is received, the table 3.2 shows the ranges of values. Activities: The only activities are sensing and taking pictures of the world. Protocols: Ultrasound: This is responsible for sensing the distance to get to an obstacle. The initiative for this would be the agent system, which indicates the time of initiation. Chamber is responsible for obtaining the images within the scene, and as for the ultrasonic sensor system depends for its initiation. Light sensors: the role of these sensors is to measure the intensity with which an object reflects light (walls, diagrams). This makes a light emitting and measuring the portion of the return is received and handled the ranges are 0 to 1023 RAW. Agent system Responsibilities: It is responsible for image processing and decision making, and sends a message to agent communication process completed? Permissions: This is the one that has full access, sensors, communication, engine and handling agents and KHEPIII in general. Activities: Making decisions based on the behavior and implements what is needed to finish the job. Task agent Responsibilities: Motor movement Volume 1, Issue 2 July-August 2012

Node Agent

- Camera

Task Agent

Motor Movement

Image Processing MAS

Coordination Agent

Figure 10 Shows the diagrams of the Agent Node (AN), Task Agent (TA), and Agent System (AS).
Node Agent Agent System Task Agent

Figure 11 Diagram of the relationship model of SMA. Protocols: The system is the initiator, resulting in a movement or departure, which can be any direction, depending on the decision taken by the agent system. Figure 9 shows the three models that form the multiagent system (MAS) using Gaia. Figure 10 presents the relationship between the agents of the multi-agent system (MAS).

Page 25

International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)

Web Site: www.ijettcs.org Email: editor@ijettcs.org, editorijettcs@gmail.com Volume 1, Issue 2, July August 2012 ISSN 2278-6856
[7] Tigli, J.Y.; Thomas, M.C., Use of multi agent systems for mobile robotics control Systems, Man, and Cybernetics, 1994. 'Humans, Information and Technology', 1994 IEEE International Conference, vol.1, no., pp.588-592 vol.1, 2-5 Oct 1994. [8] R. Liscano, R. E. Fayek, A. Manz, E. R. Stuck and J.-Y. Tigli, A Blackboard Approach to The Integration of Reactive and Strategic Reasoning for Mobile Robot Control, publi dans la revue IEEE Expert (special issue on Intelligent Robotics Systems), Volume 10, Issue 2, Pages: 24 - 36, Avril 1995, ISSN: 0885-9000. [9] Shuzhi Sam Ge and Frank L. Lewis, Autonomous Mobile Robots: Sensing, Control, Decision, Making and Applications Published in 2006 by CRC Press [10] V.Matellan, D.Borrajo ABC2 an Agenda Based Multi-Agent Model for Robots Control and Cooperation, Journal of Intelligent and Robotic Systems 32: 93114, 2001. [11] A.Drogoul, A.Collinot Applying an AgentOriented Methodology to the Design of Artificial Organizations: A Case Study in Robotic Soccer, Autonomous Agents and Multi-Agent Systems, 1, 113]129 1998. [12] O. Michel Webots: Professional mobile robot simulation. J Adv Robotics Systems 1(1):3942. AUTHORS
Said Benaissa received his Degree in High Education Deepened in Telecommunication Networks in 2008, from the National Institute of Science and Technology, Marrakech. In 2009 he joined the System Architecture Team of the ENSEM School, Casablanca, Morocco. His current main research interests embedded Systems Based on Multi-agents Systems.
2 1

Figure 11 Sample of implementation of agent

3. CONCLUSION AND FURTHER WORK

We have presented different uses of Multi Agent Systems and more particularly in the mobile robot field. We have seen that a lot of different approaches can be undertaken to design a mobile robot as a multi-agent systems and it seems very challenging to find a common representation to describe and benchmark these different approaches. In this paper we explored the possibility to describe a robot mobile such as Multi Agent Systems by decoupling a mobile robot into separately either virtual or physical autonomous entities. Following this stage of our research, the next step will be to focus separately on each of the above agents to provide more in-depth outline and examine more closely these features.

References
[1] S.Degallier, L.Righetti, S.Gay Toward simple control for complex, autonomous robotic applications: combining discrete and rhythmic motor primitives, 4 May 2011. [2] V. Vigneron, R.Kallel, H.Maaref, Un algorithme pour la planification de trajectoire bas sur le calcul ensembliste, Jan 2008, IBISC CNRS FRE, France. [3] Rafael H. Bordini, M.Dastani, Amal El Fallah Seghrouchni, Multi-Agent Programming: Languages, Tools and Applications, United Kingdom, 2009. [4] E.E.Parker and F.E.Schneider, Multi-Robot Systems. From Swarms to Intelligent Automata, Volume III, Knoxville, TN, U.S.A, 2005. [5] A.J.Ijspeert, A.Martinoli, A. Billard, M.Gambardella [6] Collaboration Through the Exploitation of Local Interactions in Autonomous Collective Robotics: The Stick Pulling Experiment Autonomous Robots 11, 149171, 2001. Volume 1, Issue 2 July-August 2012

Hicham Medromi received his PhD in engineering science from the Sophia Antipolis University in 1996, Nice, France. He is responsible of the system architecture team of the ENSEM Hassan II University, Casablanca, Morocco. His current main research interests concern Control Architecture of Mobile Systems Based on Multi Agents Systems. Since 2003 he is a full professor for automatic productic and computer sciences at the ENSEM School, Hassan II University, Casablanca.
3

Saidia Tallal: she is a full professor for computer sciences at the ENSEM School, Hassan II University, and Casablanca. Her actual main research interests concern Intrusion Detection System Based on Multi agents Systems.

Page 26