Coordination and control of multi agent systems

CONTROL & COORDINATION OF MULTI-AGENT SYSTEMS CHENG-HENG FUA B. Eng (Hons.), National University of Singapore A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY NUS GRADUATE SCHOOL FOR INTEGRATIVE SCIENCES & ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2008 Acknowledgments One phase of my life ends and another begins. I look upon the road before me with hope and with excitement, and as I face the future that is ahead, I look back upon the way I have come, and am comforted by the fact that I am not alone; that I am blessed with the great fortune of having the companionship of several distinct individuals. Like Frodo Baggins and the fellowship of the Ring, or Dorothy and her companions in the land of Oz, my own adventures would not have been possible without their support, encouragement, guidance and wisdom, and I would like to take this opportunity to express my gratitude to all of them. To my thesis supervisor, Professor Shuzhi Sam Ge, for his inspiring presence, for his constant, patient guidance, and for his selfless sharing of experiences in all things research and more. Thanks also go to Professor Khiang Wee Lim, my thesis cosupervisor, for his guidance and help on all matters concerning my research despite his busy schedule. I would also like to thank Dr Javier-Ibanez Guzman, for his insightful advice and guidance on shaping my research direction and goals. To my friends. I have been extremely fortunate to have worked with many, many brilliant people during my study, and who have always been willing and generous with their time and friendship. Special thanks to Mr Keng Peng Tee and Mr Pey Yuen Tao, my fellow intrepid adventurers, for the endless hours of discussions and brainstorming that are always filled with creativity, inspiration and crazy ideas, for their moral support and for always being there to help. To Dr Xuecheng Lai, Dr Zhuping Wang and Dr Feng Guan, for their friendship, help and guidance since the day I first joined the research team. Thanks also go to Professor Khac Duc Do for his patience, guidance and wonderful advice in my research on formation control. I would also like to thank ii Mr Chenguang Yang, Ms Chen Wang, Mr Hooman Samani, and many other fellow colleagues and researchers at the Social Robotics Laboratory for their help and friendship. To the pillars of my life – my family – I would never be where I am today without their unquestioning trust, support and encouragement. They have always been there for me, stood by me through the good times and the bad, and have always given me their full support in whatever choices I made. To my inner circle of friends – Mr Konghui Kay, Mr Khiam Boon Png, Mr Keng Chuan Ong and Mr Thian Khoon Ng – who have shown me what true friendship is, for their unwavering friendship and moral support, and for always being there in times of need. Finally, I am immensely grateful to the Agency of Science, Technology and Research (A*STAR), as well as the NUS Graduate School of Integrative Sciences and Engineering (NGS), for their funding and support, without which, this great adventure might never have taken place. iii Abstract This thesis considers in detail the technical issues associated with the effective control and coordination of Multi-Agent Systems (MAS), with particular emphasis on techniques that increase the robustness of the team of physical agents, when subjected to uncertainties, such as malfunctions and imperfect communications. For physical agent teams which are to be deployed within uncertain and dynamic environments, it is important for the team to continue functioning even in the event of unforeseen disruptions such as single-agent breakdowns and spurious communication losses, and continue being driven toward its collective goals. Such abilities are crucial for the success of autonomous teams, and would constitute the main motivation for the work presented in this thesis. The thesis is organized according to the two main decision making levels where coordination between members within an agent team can occur, namely on the Macroand Micro-levels of decision making, following the general flow of decisions, from mission specification, assignment, to actual task accomplishment. Agent cooperation on the macro-level concerns a more general management protocol that, in combination with a planning and representation framework, manages the resources (i.e. robots) and tries to arrive at a suitable distribution of tasks to either individual robots or sub-teams of robots. This level of decision making focuses on mission and task representations, and team organization algorithms. Actual task accomplishment by robot sub-teams require further, more explicit, cooperation between individual members, and this falls into the realm of micro-level coordination. Such forms of coordination is investigated in the context of representing and cooperative accomplishment of multi-agent formations. Under this framework, and with the above objectives in mind, the technical coniv tributions of this thesis provide a step towards increasing the robustness of physical agent systems operating within dynamic, uncertain operating domains. The proposed solutions include: (i) A general mission/ task representation framework based on the concept of basis tasks, that is amenable to analysis, for the efficient portrayal, subdivision, and allocation of tasks to agent teams on-the-fly. (ii) A robust, cooperative task allocation scheme, the Cooperative Back-Off Scheme (COBOS), for instantaneous task (re)distribution between spatially distant task locations which are subjected to limited communications. (iii) A representation framework for task allocation schedules over an extended time period, based on the transformation of task schedules into an agent-formation space, where stable, self-organizing, convergence algorithms are introduced to form a dynamic time-extended allocation. (iv) An efficient formation representation scheme, the Q-structure, to facilitate scalable and flexible agent formations. The Q-structure allows the representation of a wide variety of formations, and coupled with a behavior based algorithms for each agent, enables decentralized, robust formation tracking with automatic scaling. (v) A decentralized and reactive potential field based method is used for stably guiding agents into formation based only on local communications, and further subdividing the decision making process into the fast and slow time scales. Theoretical analysis have also been carried out to verify the convergence properties of the navigation controls. To further improve performance in scenarios involving limited communications, methods for dynamically adapting the short term Q-structure representation are also used. v Nomenclature R Rn×m x ˆ x(f ) g(f ) (·) x (f ) T(f21 ) Rjk ⊗ |X| ri xk |X| nj TSM(j) TSMij (j) TSM− (j) B, B bi ( i ) nb L Ti Li φi Ti,k Ti the field of real numbers; the set of n × m-dimensional real matrices; the unit vector of a vector x; a vector x expressed in the frame f . Taken to be in the world frame if unspecified; a function g(·), expressed in the frame f ; the Euclidean norm of a vector x; the transformation matrix from frame f1 to f2 ; the jk-th element of a matrix R; the vector product operator; the cardinality of a set X; the robot i; the k-th element of a vector X ∈ Rn ; the number of elements in a set X; the number of robots rj has communication links with; the Task Utility Matrix compiled by rj ; the (i, j)-th element of TSM(j); a sub-matrix made up of selected rows and columns of TSM(j); the set, or vector form, of Basis Tasks; (bi ∈ B) the i-th basis task with the list of arguments it accepts, i ; the number of basis tasks in B; the arguments accepted by basis tasks in B; the i-th macro task; the physical location associated with Ti ; the number of robots required by Ti ; the k-th sub-macro task of Ti ; the Task Specification Matrix of Ti ; vi Ti,dp |Ti | |Ti,j | Tp T atv T ach Tls nls natv Aj Trj W N nsn Sj tdd,j tsd,j td,j Aj ajk njr RSi RW Ijk U Ci F, FN Q G Qj Vj Sj Cj Ej Oj χi (t) Ej qi , qtg,i , qt vi , vtg,i , vt κi , κtg,i , κt the set of tasks that must be completed before Ti can start; the number of sub-macro tasks in Ti ; the number of basis tasks in Ti,j ; the ordered set of high priority tasks; the ordered set of active, but not high priority tasks; the ordered set of archived tasks; the ordered set of all the tasks given to the robots; the number of tasks given to the robots; the number of tasks in a subnetwork that can be considered, given the number of robots in that subnetwork, and the total number of robots required for each of these tasks; the vector containing the utility level rj has with each of the nb basis tasks; the task rj is currently performing; the physical workspace of the team; the set of disjoint subspaces in W ; the number of disjoint networks in W ; (Znb ×1 ) the Task Success Matrix of rj . the time at which task Tj must be completed, the task due date; the earliest start time for task Tj ; the approximate task duration for task Tj ; the set of agents representing the virtual-agent equivalent of task Tj ; the agent associated with task Tj on robot rk ; the number of robots that can service a task Tj ; the Roam-Space on a robot ri ; the Roam-World; the area of influence of an agent ajk ; the set of uncovered points in RSi ; the desired formation consisting of N robots; the set of all the queues in a formation F; the set of all the formation vertices; the j-th queue in the set Q; a list of (either one or two) formation vertices that influences Qj ; the set of points describing the shape of Qj ; the capacity ∈ [0, 1] of Qj ; the encapsulating region of Qj ; the set of functions that describe the orientation of agents along Qj is the queue status of ri at t; excess length (the number of excess robots in Qj ); position of robot ri , its target, and team’s target respectively; velocity of robot ri , its target, and team’s target respectively; (topside-vector) vector normal to the plane of robot ri , its target, and team’s target respectively; vii q(vj),i,nr x(f ),i,t j,i,nr Ntot Nv , Nq ρsf ρadp ρ0 Rmax Ract nR,t (φ, θ) U b , Fb Fb ab δ, δa cig,on vmax , ωmax the vector from a robot ri at q(vj),i to the nearest point on its queue Qj at q(vj),nr ; the relative position (x = q), velocity (x = v) or topside-vector (x = κ) of target w.r.t. ri in frame f ; the shortest distance between ri and a queue Qj ; the number of robots currently in the team; the number of formation vertices and queues respectively; the safety distance between ri and an obstacle; the distance a deformed queue is from an obstacle; the influence range of an obstacle; the maximum range of ri ’s range sensor; (≤ Rmax ), the active range of the Instant Goal behavior; the unit vector of the ray from the range readings that is closest to qˆ(ri),i,t ; a pair representing the direction of an arbitrary point in the frame of ri . the potential function and force derived for a behavior b; the magnitude of the force Fb ; the weighting parameter for a behavior b; the average distance of robots from their queues and deformed queues (if applicable) respectively; ON(1)/OFF(0) status of Instant Goal Behavior; the maximum speed and turnrate of a robot respectively. viii Contents Acknowledgments ii Abstract iv Nomenclature vi Table of Contents xi List of Figures xi List of Tables xiv Introduction 1.1 Motivation of Research: Multi-Agent Coordination . 1.2 Macro-Level Planning & Inter-Task Coordination . . 1.2.1 Task Representation & Short Term Allocation 1.2.2 Plan Representation & Long Term Allocation 1.3 Micro-Level Coordination . . . . . . . . . . . . . . 1.3.1 Formation Representation & Control . . . . . 1.4 Contributions . . . . . . . . . . . . . . . . . . . . . 1.5 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 10 10 14 15 Task Representation and Short Term Allocation 2.1 The Cooperative BackOff Adaptive Scheme . . . . . . . . . . . . . . . 2.1.1 Disjoint Broadcast Networks . . . . . . . . . . . . . . . . . . . 2.1.2 Formal Description of Tasks . . . . . . . . . . . . . . . . . . . 2.1.3 Task Suitability Matrices (TSM) . . . . . . . . . . . . . . . . . 2.1.4 Fault Tolerance and Coping with Uncertain Task Specifications 2.1.5 Adaptation of Internal Robot Model . . . . . . . . . . . . . . . 2.2 Task Prioritization and Allocation . . . . . . . . . . . . . . . . . . . . 2.3 Analysis and Comparisons . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Domain of Operation . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Communication Complexity . . . . . . . . . . . . . . . . . . . 2.3.3 Computation Complexity . . . . . . . . . . . . . . . . . . . . . 2.3.4 Quality of Solutions . . . . . . . . . . . . . . . . . . . . . . . 2.4 Simulation Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . 17 18 19 20 26 28 30 32 38 39 41 41 42 44 ix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 45 50 51 Plan Representation and Long Term Allocation 3.1 Decentralized Task Scheduling . . . . . . . . . . . . . . . . 3.1.1 Self-Organizing Schedules . . . . . . . . . . . . . . 3.2 Agent Dynamics and Behavior . . . . . . . . . . . . . . . . 3.3 Simulation Experiments . . . . . . . . . . . . . . . . . . . . 3.3.1 Convergence of Agents with Known Task Durations 3.3.2 Tasks with High Degree of Uncertainty . . . . . . . 3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 55 55 60 65 65 67 69 Formation Representation with Q-Structures 4.1 Queues and Artificial Potential Trenches . . . . 4.1.1 Assumptions . . . . . . . . . . . . . . 4.1.2 Formations and Queues . . . . . . . . . 4.1.3 Changing Queues . . . . . . . . . . . . 4.1.4 Potential Trench Functions . . . . . . . 4.2 Robot Behaviors . . . . . . . . . . . . . . . . 4.2.1 Target Tracking . . . . . . . . . . . . . 4.2.2 Instant Goal Behavior . . . . . . . . . 4.2.3 Obstacle Avoidance . . . . . . . . . . . 4.2.4 Overall Robot Behavior . . . . . . . . 4.3 Analysis of Parameter Values . . . . . . . . . . 4.4 Simulation Experiments . . . . . . . . . . . . . 4.4.1 Convergence to Formations and Scaling 4.4.2 Maneuvers in Confined Spaces . . . . . 4.4.3 Reaction of Formations to obstacles . . 4.4.4 Disruption of Wireless Communications 4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 71 71 72 75 77 80 80 82 85 86 86 89 91 93 97 97 102 Q-Structures and Formation Convergence with Limited Communication 5.1 Formation Representation and Dynamic Target Determination . . . . . 5.1.1 Division of Information Flow . . . . . . . . . . . . . . . . . . 5.1.2 Properties of the Q-structure . . . . . . . . . . . . . . . . . . . 5.1.3 Determination of Target on Queue . . . . . . . . . . . . . . . . 5.2 Navigation of Robots to Positions in Formations . . . . . . . . . . . . . 5.3 Simulation Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Formation Convergence and Scaling . . . . . . . . . . . . . . . 5.3.2 Moving formations . . . . . . . . . . . . . . . . . . . . . . . . 5.3.3 Changing Formations . . . . . . . . . . . . . . . . . . . . . . . 5.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 104 105 106 110 111 121 121 124 124 126 127 2.5 2.4.1 Tasks and Mission Statement . . . . . . . . . . 2.4.2 Tasks in Connected Communication Networks 2.4.3 Tasks in Disjoint Communication Networks . . Summary . . . . . . . . . . . . . . . . . . . . . . . . x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chapter Conclusions and Recommendations The work presented in this thesis focused on the development of effective methodologies in synthesizing successful and effective behaviors for multi-robot collaboration on two main levels – the macro and micro levels. In this chapter, the results of the research work described in the previous chapters are summarized and the major contributions of this work are reviewed. Suggestions for future work are also presented. 7.1 Summary and Contributions This thesis has covered areas pertaining to both the macro and micro levels of decision making in multi-robot teams with the main aim of synthesizing behaviors for coping with uncertain operating conditions. The approach that has been adopted for each section of the thesis consider the problems from mainly the representational and behavioral generation points of view. That is to say, in each section, the thesis had first considered effective representations of information, such as tasks and formations, suitable for the domains of operation, and follows with the design of suitable, complementary robot behaviors. To this end, the thesis has attempted to develop task/plan specification frameworks upon which task allocation can be made. A concise formation representation scheme has also been developed. Behavioral algorithms have also been developed to cater to uncertain operating conditions and communications/sensing limitations. The key results presented in this thesis are as follows: 161 7.1. Summary and Contributions • Task Representation and Short Term Allocation (Chapter 2): A matrix-based approach has been proposed for specifying tasks (accounting for task dependen- cies and alternative methods of performing tasks) using basic task-achieving behaviors. This representation improves the portability of tasks across different robot teams and allows robots to make use of task success/failure histories to detect and adapt to device imperfections and malfunctions. In addition, the Cooperative Back-Off Adaptive Scheme (COBOS) is proposed to achieve fault-tolerant task allocation. It is designed to improve robot autonomy in ST-MR-IA domains with uncertain task specifications, and with disjoint communication networks. • Plan Representation and Long Term Allocation (Chapter 3): A self-organizing agent-based framework, with the use of a Dandelion network/graph that defines agent neighborhood relationships has been proposed for representing tasks and plan allocations. The proposed framework allows decentralized schedule formation and equalization of work load on robots. In addition, algorithms that govern agent interactions and produce clustering behavior have been proposed. Under the influence of these algorithms, the agents are able to self-organize to produce a feasible and flexible schedule (respecting constraints), while equalizing work loads (as far as possible) between robots based on the latest expected completion date of all the tasks each robot is handling. • Formation Representation with Q-Structures (Chapter 4): The concept of queues, instead of nodes, has been proposed as a novel and flexible methodology to define and support a large variety of formations. Based on the proposed representation, a decentralized redistribution algorithm has been designed for the robot team to redistribute themselves dynamically amongst queues in response to changes in the formation or in the number of robots in the team. Furthermore, conventional methods using potential fields, which attract robots to predetermined points in order to form the desired pattern, limit the formations scalability. In an attempt to address this drawback, the chapter has presented a generalization of conventional potential field methods to lines through the use of the 162 7.1. Summary and Contributions artificial potential trench, each associated with a queue. With potential trenches, each robot was attracted to, and moves along the bottom of the valley created by the potential trench, thereby automatically distributing themselves along the trench. • Q-Structures and Formation Convergence with Limited Communications (Chapter 5): The Q-structure has been extended by considering finite communi- cation ranges, and separating the decision making process into two time scales - a fast time scale for reactive decision making based only on local communications, and a slower time scale which allows less time critical information to propagate through a weakly connected network. Persistent global communications is not required, reducing overall communications load. It also permits intermittent information losses as information is collected over a longer time. In addition, a rigorous proof of convergence for a decentralized control law that guides robots into formations represented by the Q-structure has been proposed. A dynamic target determination algorithm has also been proposed to incrementally guide robots into their queues. This proposed algorithm improves upon the original scheme by separating robot decisions regarding positions on queues from reactive inter-robot repulsive forces, to obtain an even distribution of robots along queues. • Adaptive Q-Structures and Formation Convergence with Motion Limitations (Chapter 6): The Q-formation scheme has been further extended into the 3-D space and orientation information has been incorporated into the representation. Furthermore, the organizational structure of the Q-structure has been exploited to explicitly segregate short term information flow in the system and to adapt the short term communication structure according to communication ranges. The limitations on the amount of direction changes each robot is capable of making at each instant has been considered, such that each robot would prefer gradual directional changes instead of abrupt turns. Constellation-agents have been proposed and are used by each robot to bias their motion to reflect such preferences. The concept of Cast-Zones and bobber-agents have been pro163 7.2. Suggestions for Future Work posed and used by each vehicle to generate suitable intermediate targets between the vehicle and their actual target on the queue. These intermediate targets are determined by the movement and convergence of the bobber-agents in their associated cast-zones. The intermediate targets acts as a more appropriate target for the vehicles by reducing the immediate need for sudden directional changes. 7.2 Suggestions for Future Work After a review of the research work, this section presents the directions that are recommended for extending the results developed in this thesis: • Mission accomplishment being the main concern in the chapters on macro-level decision making, the proposed task allocation methodologies using both COBOS and the self-organizing dandelion framework made the common assumption that there is no cost involved in task changeovers. That is, the changeover cost – for example in terms of time, additional distance traveled, adaptation of capabilities – of having one robot (or a sub-team) take over a task from another has not been considered. This issue, however, can have profound implications on the efficiency of a team when the mission is time critical. One possible solution is the incorporation of such costs as a variable when determining task suitability. • The main domains of operation under consideration in the thesis is that of single/multirobot tasks with either instantaneous or time-extended allocation. The domain where multi-tasking robots are present (i.e., the Multi-Task Robots domain) have not been considered. This can also refer to domains where several tasks within a mission have congruent or overlapping objectives, and which makes use of the same capabilities within a robot such that the accomplishment of one task can possibly imply the concurrent success of another. For instance, such tasks can take the form of a target detection task and an area surveillance/scouting task. These two tasks have differing objectives, but have the potential of being achieved concurrently, a potential that should be exploited as much as possible to make effi164 7.2. Suggestions for Future Work cient use of available resources. In these domains, concurrent allocations must be performed. Under the existing task allocation framework, which has the advantage of allowing easy identification of resources required per task, an additional module could possibly be included for identifying and merging such tasks. • Task suitability has been defined in terms of the match between task requirements and robot capabilities/resources. The evaluation of team suitability presented in this thesis assumes that task suitability is additive due to highly uncertain operating conditions where information is scarce. However, when information (of other robots) is available, team efficiency can be improved through the consideration of “team synergy” when computing the suitability of a sub-team for a certain Multi-Robot task. One possible approach could be the development of “cliques” within the team based on a decentralized a priori evaluation of individual robot capabilities by each robot within the team. The current algorithm in COBOS relates task successes/failures to individual accomplishments, and not the team involved. This promotes a self-centered view within each robot and limits the development of collaborative experience. Therefore, another approach would be the integration of an “experience” component in the evaluation of task suitability that evolves in real-time, and is based on task successes and failures, in relation to the sub-teams involved. • In our consideration of multi-robot formation schemes, each robot has been rep- resented with a point mass. The effect of movement constraints have been briefly considered in the last chapter. An interesting extension will be to synthesize formation control laws and algorithms that directly considers the dynamic motion constraints of each robot to generate smooth trajectories. In this aspect, the formation control techniques presented in this thesis could be merged with lower level path planning approaches to produce feasible paths for more sophisticated systems. • Based on the Q-structure, the stable control laws described in the later chapters 165 7.2. Suggestions for Future Work of the thesis considers an environment with point sized obstacles, which mainly includes the other robots within the team but can be easily extended through a slight modification of the obstacle repulsion function Uob . One limitation of this is that the problem of local minima created if large obstacles are present in the environment has not been considered. This has important implications on mobile systems in dynamic environments, and should be a natural extension of the work presented in this thesis. A possible solution is the integration of more sophisticated path planning methods (e.g., the Instant-Goal approach in [94]) into the system to produce intermediate waypoints for the team, and the stability of the system re-evaluated based on a changing set of targets. 166 Bibliography [1] B. P. Gerkey and M. J. Matari´c, “A Formal Analysis and Taxonomy of Task Allocation in Multi-Robot Systems,” International Journal of Robotics Research, vol. 23, no. 9, pp. 939–954, September 2004. [2] T. Balch and L. E. Parker, Eds., Robot Teams: From Diversity to Polymorphism. Natick, MA: A K Peters Ltd., 2001. [3] K. P. Sycara, “Multiagent systems,” AI Magazine, vol. 10, no. 2, pp. 79–93, 1998. [4] B. P. Gerkey and M. J. Matari´c, “Sold!: Auction Methods for Multirobot Coordination,” IEEE Transactions on Robotics and Automation, vol. 18, no. 5, pp. 758–768, October 2002. [5] A. Ortony, D. A. Norman, and W. Revelle, “Affect and Proto-Affect in Effective Functioning,” In J.M. Fellous and M.A. Arbib, Who needs emotions: The brain meets the machine, 2005. [6] L. E. Parker, “ALLIANCE: An Architecture for Fault Tolerant Multirobot Cooperation,” IEEE Transactions on Robotics and Automation, vol. 14, no. 2, pp. 220–240, April 1998. [7] C. Fua, S. S. Ge, and K. W. Lim, “BOAs: Back-Off Adaptive Scheme for Task Allocation with Fault Tolerance and Uncertainty Management,” Proc. International Symposium on Intelligent Control, pp. 162–167, September 2004. [8] C. Fua and S. S. Ge, “COBOS: Cooperative Back-Off Adaptive Scheme for MultiRobot Task Allocation,” IEEE Transactions on Robotics, vol. 21, no. 6, pp. 1168– 1178, December 2005. [9] L. E. Parker, “L-ALLIANCE: Task-oriented multirobot learning in behavior-based systems,” Advanced Robotics, vol. 11, no. 4, pp. 305–322, 1997. [10] B. B. Werger and M. J. Matari´c, “Broadcast of Local Eligibility for Multi Target Observation,” in Distributed Autonomous Robotic Systems 4, L. E. Parker, G. Bekey, and J. Barhen, Eds. Springer, 2001, pp. 347–356. [11] M. Tambe and W. Zhang, “Towards Flexible Teamwork in Persistent Teams: Extended Report,” Autonomous Agents and Multiagent Systems, vol. 3, no. 2, pp. 159–183, June 2000. 167 Bibliography [12] N. Miyata, J. Ota, T. Arai, and H. Asama, “Cooperative Transport by Multiple Mobile Robots in Unknown Static Environments Associated with Real-Time Task Assignment,” IEEE Transactions on Robotics and Automation, vol. 18, no. 5, pp. 769–780, October 2002. [13] S. C. Botelho and R. Alami, “M+: A Scheme for Multi-Robot Cooperation through Negotiated Task Allocation and Achievement,” Proc. International Conference on Robotics and Automation, pp. 1234–1239, May 1999. [14] O. Shehory and S. Kraus, “Methods for Task Allocation via Agent Coalition Formation,” Artificial Intelligence, vol. 101, no. 1-2, pp. 165–200, May 1998. [15] R. Zlot and A. Stentz, “Market-Based Multi-Robot Coordination using Task Abstraction,” 4th Int. Conf. Field and Service Robotics, July 2003. [16] N. Kalra, T. Stentz, and D. Fergusen, “Hoplites: A Market Framework for Complex Tight Coordination in Multi-Agent Teams,” Robotics Institute, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, CMU-RI-TR-04-41, August 2004. [17] K. M. Sim and E. Wong, “Toward market-driven agents for electronic auction,” IEEE Transactions on Systems, Man and Cybernetics - Part A, vol. 31, no. 6, pp. 474–484, November 2001. [18] L. Chaimowicz, M. F. M. Campos, and V. Kumar, “Dynamic role assignment for cooperative robots,” Proc. International Conference on Robotics and Automation, pp. 293–298, May 2002. [19] L. Chaimowicz, V. Kumar, and M. F. M. Campos, “A Paradigm for Dynamic Coordination of Multiple Robots,” Autonomous Robots, vol. 17, pp. 7–21, 2004. [20] R. Emery, K. Sikorski, and T. Balch, “Protocols for collaboration, coordination and dynamic role assignment in a robot team,” Proc. IEEE Int. Conf. on Robotics and Automation, pp. 3008–3015, May 2002. [21] R. Zlot, A. Stentz, M. B. Dias, and S. Thayer, “Multi Robot Exploration Controlled by a Market Economy,” Proc. IEEE International Conference on Robotics and Automation, pp. 3016–3023, May 2002. [22] M. B. Dias, M. Zinck, R. Zlot, and A. Stentz, “Robust Multirobot Coordination in Dynamic Environments,” Proc. IEEE Int. Conf. Robotics and Automation, pp. 3435–3442, April 2004. [23] S. Kreipl and M. Pinedo, “Planning and Scheduling in Supply Chains: An Overview of Issues in Practice,” Production and Operations Management, vol. 13, no. 1, pp. 77–92, 2004. 168 Bibliography [24] M. T. Jenson, “Generating Robust and Flexible Job Shop Schedules using Genetic Algorithms,” IEEE Trans. Evolutionary Computation, vol. 7, no. 3, pp. 275–288, June 2003. [25] E. Hart, P. Ross, and J. Nelson, “Producing Robust Schedules via an Artificial Immune System,” Proc. IEEE Int. Conf. Evolutionary Computation, IEEE World Congress on Computational Intelligence, pp. 464–469, May 1998. [26] C. Artigues, P. Michelon, and S. Reusser, “Insertion Techniques for Static and Dynamic Resource-Constrained Project Scheduling,” European Journal of Operational Research, vol. 149, pp. 249–267, 2003. [27] T. K. Tsukada and K. G. Shin, “Distributed Tool Sharing in Flexible Manufacturing Systems,” IEEE Trans. Robotics and Automation, vol. 14, no. 3, pp. 379–389, June 1998. [28] Y. L. Qu´er´e, M. Sevaux, C. Tahon, and D. Trentesaux, “Reactive Scheduling of Complex System Maintenance in a Cooperative Environment with Communication Times,” IEEE Tran. Systems, Man, and Cybernetics - Part C, vol. 33, no. 2, pp. 225–234, May 2003. [29] H. Toba, “Segment-Based Approach for Real-Time Reactive Rescheduling for Automatic Manufacturing Control,” IEEE Tran. Semiconductor Manufacturing, vol. 13, no. 3, pp. 264–272, August 2000. [30] B. Hamidzadeh, L. Y. Kit, and D. J. Lilja, “Dynamic Task Scheduling using Online Optimization,” IEEE Tran. Parallel and Distributed Systems, vol. 11, no. 11, pp. 1151–1163, November 2000. [31] D. Montana, M. Brinn, S. Moore, and G. Bidwell, “Genetic algorithms for complex, real-time scheduling,” Proc. IEEE Int. Conf. Systems, Man, and Cybernetics, vol. 3, pp. 2213–2218, October 1998. [32] P. Sousa and C. Ramos, “A Distributed Architecture and Negotiation Protocol for Scheduling in Manufacturing Systems,” Computers in Industry, vol. 38, pp. 103– 113, 1999. [33] G. E. Vieira, J. W. Herrmann, and E. Lin, “Rescheduling Manufacturing Systems: A Framework of Strategies, Policies, and Methods,” Journal of Scheduling, vol. 6, pp. 39–62, 2003. [34] M. Fabiunke and G. Kock, “A Connectionist Method to Solve Job Shop Problems,” Proc. 2nd Int. Conf. Intelligent Processing and Manufacturing of Materials, vol. 1, pp. 655–660, July 1999. [35] R. Olfati-Saber, W. B. Dunbar, and R. M. Murray, “Cooperative control of multivehicle systems using cost graphs and optimization,” Proc. American Control Conference, vol. 3, pp. 2217–2222, June 2003. 169 Bibliography [36] N. E. Leonard and E. Fiorelli, “Virtual Leaders, Artificial Potentials and Coordinated Control of Groups,” Proc. IEEE Conference on Decision and Control, vol. 3, pp. 2968–2973, December 2001. [37] T. Balch and R. C. Arkin, “Behavior-Based Formation Control for Multirobot Teams,” IEEE Transactions on Robotics and Automation, vol. 14, pp. 926–939, 1998. [38] A. K. Das, R. Fierro, V. Kumar, J. P. Ostrowski, J. Spletzer, and C. J. Taylor, “A Vision Based Formation Control Framework,” IEEE Transactions on Robotics and Automation, vol. 18, pp. 813–825, 2002. [39] J. P. Desai, J. P. Ostrowski, and V. Kumar, “Modeling and Control of Formations of Nonholonomic Mobile Robots,” IEEE Transactions on Robotics and Automation, vol. 17, pp. 905–908, 2001. [40] M. Egerstedt and X. Hu, “Formation constrained multi−agent control,” IEEE Transactions on Robotics and Automation, vol. 17, no. 6, pp. 947–951, 2001. [41] W. Ren and R. W. Beard, “A Decentralized scheme for Spacecraft Formation Flying via the Virtual Structure Approach,” AIAA Journal of Guidance, Control, and Dynamics, vol. 27, no. 1, pp. 73–82, January 2004. [42] M. Tillerson, L. Breger, and J. P. How, “Distributed coordination and control of formation flying spacecraft,” Proc. American Control Conference, vol. 2, pp. 1740–1745, June 2003. [43] X. Yun, G. Alptekin, and O. Albayrak, “Line and Circle Formation of Distributed Physical Mobile Robots,” Journal of Robotic Systems, vol. 14, no. 2, p. 6376, February 1997. [44] T. Eren, P. N. Belhumeur, and A. S. Morse, “Closing Ranks in Vehicle Formations based on Rigidity,” Proc. IEEE Conference on Decision and Control, pp. 2959– 2964, December 2002. [45] T. Balch and M. Hybinette, “Social Potentials for Scalable Multi Robot Formations,” Proc. IEEE International Conference on Robotics and Automation, pp. 73–80, April 2000. [46] S. Carpin and L. E. Parker, “Cooperative leader following in a distributed multirobot system,” Proc. International Conference on Robotics and Automation, pp. 2994–3001, May 2002. ˇ [47] P. Kostelnik, M. Samulka, and M. J´anoˇs´ık, “Scalable Multi-Robot Formations using Local Sensing and Communication,” Third International Workshop on Robot Motion and Control, pp. 319–324, November 2002. 170 Bibliography [48] W. Kang and N. Xi, “Control and adaptation of multi vehicle formations,” Proc. International Conference on Intelligent Robots and Systems, pp. 1155–1160, 1999. [49] W. Kang, N. Xi, Y. Zhao, J. Tan, and Y. Wang, “Formation control of multiple autonomous vehicles: Theory and experimentation,” IFAC 15th Triennial World Congress, pp. 1155–1160, 2002. [50] D. M. Stipanovi´c, G. ˙Inalhan, R. Teo, and C. J. Tomlin, “Decentralized Overlapping Control of a Formation of Unmanned Aerial Vehicles,” Automatica, vol. 40, no. 8, pp. 1285–1296, August 2004. [51] P. Song and V. Kumar, “A Potential Field Based Approach to Multi Robot Manipulation,” Proc. IEEE International Conference on Robotics and Automation, pp. 1217–1222, May 2002. [52] H. G. Tanner, G. J. Pappas, and V. Kumar, “Leader-to-Formation Stability,” IEEE Trans. on Robotics and Automation, vol. 20, no. 3, pp. 443–455, June 2004. [53] E. Rimon and D. E. Koditschek, “Exact Robot Navigation Using Artificial Potential Functions,” IEEE Trans. Robotics and Automation, vol. 8, no. 5, pp. 501–518, October 1992. [54] H. Tanner and A. Kumar, “Towards Decentralization of Multi-Robot Navigation Functions,” Proc. IEEE Int. Conf. Robotics and Automation, pp. 4143–4148, April 2005. ¨ [55] P. Ogren and N. E. Leonard, “Obstacle Avoidance in Formation,” IEEE Int. Conf. Robotics and Automation, pp. 2492–2497, September 2003. [56] Y. Liu, K. M. Passino, and M. Polycarpou, “Stability Analysis of OneDimensional Asynchronous Swarms,” IEEE Trans. Automatic Control, vol. 48, no. 10, pp. 1848–1854, October 2003. [57] A. Khoo and I. D. Horswill, “An efficient coordination architecture for autonomous robot teams,” Proc. Internation Conference on Robotics and Automation, pp. 287–292, May 2002. [58] O. B. Bayazit, J.-M. Lien, and N. M. Amato, “Roadmap-Based Flocking for Complex Environments,” Proc. 10th Pacific Conference on Computer Graphics and Applications, pp. 104–113, October 2002. [59] J. Fredslund and M. J. Matari´c, “Robot Formations Using Only Local Sensing and Control,” IEEE International Symposium on Computational Intelligence in Robotics and Automation, pp. 308–313, July, August 2001. [60] L. E. Parker, “The Effect of Heterogeneity in Teams of 100+ Mobile Robots,” Multi-Robot Systems Volume II: From Swarms to Intelligent Automata, Kluwer, pp. 205–215, 2003. 171 Bibliography [61] C. Reynolds, “Flocks, Herds and schools: A distributed behavioral model,” ACM SIGGRAPH Computer Graphics, Proc. 14th Conference on Computer Graphics and Interactive Techniques, vol. 21, no. 4, pp. 25–34, August 1987. [62] S. S. Ge and Y. J. Cui, “Dynamic Motion Planning for Mobile Robots Using Potential Field Method,” Autonomous Robots, vol. 13, pp. 207–222, 2002. [63] S. S. Ge and C. Fua, “Queues and Artificial Potential Trenches for Multi-Robot Formations,” IEEE Trans. Robotics, vol. 21, no. 3, pp. 646–656, August 2005. [64] ——, “Complete Multi-Robot Coverage of Unknown Environments with Minimum Repeated Coverage,” Proc. International Conf. Robotics and Automation, pp. 727–732, April 2005. [65] M. Pinedo, Scheduling: Theory, Algorithms and Systems, 2nd ed. Upper Saddle, N.J.: Prentice Hall, 2002. [66] D. Gauthier, P. Freedman, G. Carayannis, and A. S. Malowany, “Interprocess Communication for Distributed Robotics,” IEEE Journal of Robotics and Automation, vol. RA-3, no. 6, pp. 493–504, December 1987. [67] S. Lee, K. C. Lee, M. H. Lee, and F. Harashima, “Integration of Mobile Vehicles for Automated Material Handling using Profibus and IEEE 802.11 Networks,” IEEE Transactions on Industrial Electronics, vol. 49, no. 3, pp. 693–701, June 2002. [68] T. Balch and R. C. Arkin, “Communication in reactive multi agent robotic systems,” Autonomous Robots, vol. 1, no. 1, pp. 27–52, 1995. [69] L. E. Parker, “The Effect of Action Recognition and Robot Awareness in Cooperative Robotic Teams,” Proc. International Conference on Intelligent Robots and Systems, vol. 1, pp. 212–219, August 1995. [70] L. Lin and Z. Zheng, “Combinatorial Bids Based Multi-Robot Task Allocation Method,” Proc. IEEE Int. Conf. Robotics and Automation, pp. 1157–1162, April 2005. [71] F. Tang and L. E. Parker, “ASyMTRe: Automated Synthesis of Multi-Robot Task Solutions through Software Reconfiguration,” Proc. IEEE Int. Conf. Robotics and Automation, pp. 1513–1520, April 2005. [72] A. Stroupe, R. Ravichandran, and T. Balch, “Value-Based Action Selection for Exploration and Dynamic Target Observation with Robot Teams,” Proc. Int. Conf. Robotics and Automation, pp. 4190–4197, April 2004. [73] K. L. Hoffman and M. W. Padberg, “Solving Airline Crew Scheduling Problems by Branch-and-Cut,” Management Science, vol. 39, no. 6, pp. 657–682, July 1993. 172 Bibliography [74] G. B. Dantzig and M. N. Thapa, Linear Programming, 1: Introduction. Springer, 1997. [75] B. P. Gerkey, R. T. Vaughan, and A. Howard, “The Player/Stage Project: Tools for Multi-Robot and Distributed Sensor Systems,” Proc. International Conference on Advanced Robotics, pp. 317–323, June-July 2003. [76] R. Olfati-Saber and R. M. Murray, “Flocking with Obstacle Avoidance: Cooperation with limited Communication in Mobile Networks,” Proc. IEEE Conference on Decision and Control, vol. 2, pp. 2022–2028, December 2003. [77] S. S. Ge and Y. J. Cui, “New Potential Functions for Mobile Robot Path Planning,” IEEE Transactions on Robotics and Automation, vol. 16, no. 5, pp. 615–620, 2000. [78] S. S. Ge, Y. J. Cui, and C. Zhang, “Instant-Goal-Driven Methods for BehaviorBased Mobile Robot Navigation,” Proc. IEEE International Symposium on Intelligent Control, pp. 269–275, 2003. [79] P. Fiorini and Z. Shiller, “Motion planning in dynamic environments using velocity obstacles,” International Journal of Robotics Research, vol. 17, no. 7, pp. 760–772, July 1998. [80] B. MacLennan, “Synthetic Ethology: An Approach to the Study of Communication,” Artificial Life II: 2nd Workshop on Synthesis and Simulation of Living Systems, vol. 11, pp. 631–658, 1991. [81] L. E. Parker, “Designing Control Laws for Cooperative Agent Teams,” Proc. International Conference on Robotics and Automation, vol. 3, pp. 582–587, 1993. [82] K. D. Do and S. S. Ge, “Nonlinear Formation Control of Mobile Agents with Collision Avoidance,” Automatica, Submitted October 2005. [83] K. D. Do, “Bounded Controllers for Formation Stabilization of Mobile Agents with Limited Sensing Ranges,” IEEE Trans. Automatic Control, vol. 52, no. 3, pp. 569–576, March 2007. [84] A. Muhammad and M. Egerstedt, “Connectivity Graphs as Models of Local Interactions,” Journal of Applied Mathematics and Computation, vol. 168, no. 1, pp. 243–269, September 2005. [85] A. Muhammad, M. Ji, and M. Egerstedt, “Applications of Connectivity Graph Processes in Networked Sensing and Control,” Workshop on Networked Embedded Sensing and Control, October 2005, notre Dame, IN, USA. [86] H. G. Tanner, A. Jadbabaie, and G. J. Pappas, “Flocking in fixed and switching networks,” IEEE Trans. Automatic Control, submitted 2004/2005. [87] M. Egerstedt, A. Muhammad, and X. Hu, “Formation Control Under Limited Sensory Range Constraints,” 10th Mediterranean Conf. Control and Automation, July 2002, lisbon, Portugal. 173 Bibliography [88] H. G. Tanner, “Flocking with Obstacle Avoidance in Switching Networks of Interconnected Vehicles,” IEEE Int. Conf. Robotics and Automation, pp. 3006–3011, April 26-May 2004, New Orleans, LA. [89] R. Olfati-Saber and R. M. Murray, “Distributed structural stabilization and tracking for formations of dynamic multi agents,” Proc. IEEE Conference on Decision and Control, pp. 209–215, December 2002. [90] H. Axelsson, A. Muhammad, and M. Egerstedt, “Autonomous Formation Switching for Multiple Mobile Robots,” IFAC Conf. Analysis and Design of Hybrid Systems, June 2003, sant-Malo, Brittany, France. [91] The Player and Stage Project, Available from http://playerstage.sourceforge.net/. [92] S. S. Ge, C.-H. Fua, K. D. Do, and K. W. Lim, “Multi-Robot Formations Based on the Queue-Formation Scheme With Limited Communications,” IEEE Trans Robotics, 2007, conditionally Accepted. [93] A. Fujimori, M. Teramoto, P. N. Nikiforuk, and M. M. Gupta, “Cooperative Collision Avoidance between Multiple Mobile Robots,” Journal of Robotic Systems, vol. 17, no. 7, pp. 347–363, 2000. [94] S. S. Ge, X. Lai, and A. A. Mamun, “Boundary Following and Globally Convergent Path Planning Using Instant Goals,” IEEE Trans. Systems, Man and Cybernetics – Part B, vol. 35, no. 2, pp. 240–254, 2005. [95] T. J. Koo and S. Sastry, “Design of a Helicopter Model Based on Approximate Linearization,” Proc. IEEE Conf. Decision and Control, pp. 3635–3640, December 1998. [96] S. Zelinski, T. J. Koo, and S. Sastry, “Hybrid System Design for Formations of Autonomous Vehicles,” Proc. IEEE Conf. Decision and Control, vol. 1, pp. 1–6, December 2003. [97] A. Pant, P. Seiler, T. J. Koo, and K. Hedrick, “Mesh Stability of Unmanned Aerial Vehicle Clusters,” Proc. American Control Conference, vol. 1, pp. 62–68, June 2001. [98] R. Olfati-Saber, “Flocking for Multi-Agent Dynamic Systems: Algorithms and Theory,” IEEE Trans. Automatic Control, vol. 51, no. 3, pp. 401–420, March 2006. 174 Appendix A Author’s Publications Journal Papers: [1] S. S. Ge and C. Fua, “Queues and Artificial Potential Trenches for MultiRobot Formations,” IEEE Transactions Robotics, vol. 21, no. 3, pp. 646656, August 2005. [2] C. Fua and S. S. Ge, “COBOS: Cooperative Back-Off Adaptive Scheme for Multi-Robot Task Allocation,” IEEE Transactions on Robotics, vol. 21, no. 6, pp. 1168-1178, December 2005. [3] C. Fua, S. S. Ge, K. D. Do, and K. W. Lim, “Q-Structure Based MultiRobot Formations With Limited Communications,” IEEE Transactions on Robotics, Vol. 23, no. 6, pp. 1160-1169, December 2007. Conference Papers: [1] S. S. Ge, C. Fua and K. W. Lim, “Multi-Robot Formations: Queues and Artificial Potential Trenches,” Proceedings of 2004 IEEE International Conference on Robotics and Automation (ICRA), pp.3345-3350, New Orleans, Louisiana, April 26 - May 1, 2004. [2] C. Fua, S.S. Ge, K. W. Lim, “BOAs: Backoff adaptive scheme for task allocation with fault tolerance and uncertainty management,” Proceedings of the 19th IEEE International Symposium on Intelligent Control (ISIC), pp. 162-167, Taipei, Taiwan, 2- September 2004. [3] S. S. Ge, C. Fua, and W. M. Liew, “Swarm Formations using the General Formation Potential Function,” Proceedings of IEEE Conference on Robotics, Automation and Mechatronics (RAM), pp. 655-660, December 2004, Singapore. 175 [4] S. S. Ge and C. Fua, “Complete Multi-Robot Coverage of Unknown Environments with Minimum Repeated Coverage,” Proceedings of IEEE International Conference on Robotics and Automation (ICRA), pp. 727-732, Barcelona, Spain, 18-22 April 2005. [5] C. Fua, S. S. Ge and K. W. Lim, “Fault Tolerant Task Scheduling for MultiRobot Teams using Self-Organizing Agents in Formation,” Proceedings of IEEE International Conference on Robotics and Automation (ICRA), p.576-581, Orlando, Florida, USA, May 15-19, 2006. [6] S. S. Ge, C. Fua, K. D. Do, and K. W. Lim, “Multi-Robot Formations based on the Queue-Formation Scheme with Limited Communications,” Proceedings of IEEE International Conference on Robotics and Automation (ICRA), pp. 2385-2390, Rome, Italy, April 10-14, 2007. [7] C. Fua, S. S. Ge, J. B. Zhang and K. W. Lim, “Adaptive Q-Structure for Agent Formations,” Proceedings of 22nd IEEE International Symposium on Intelligent Control (ISIC), IEEE Multi-Conference on Systems and Control (MSC), pp. 518-523, Singapore, October 1-3, 2007. 176 [...]... light of the growing autonomy of singular mobile robots and the great potential of collaborative teams, this thesis focuses on the efficient coordination of embodied robots, with the main aim of improving the autonomy of robot teams operating in dynamic 1 1.1 Motivation of Research: Multi- Agent Coordination and unknown environments Embodied robot systems are significantly different from typical multi- agent. .. outline of this thesis is also presented 1.1 Motivation of Research: Multi- Agent Coordination Multi- Agent Systems, consisting of huge numbers of interacting agents, linked together in complex networks, are becoming increasingly prevalent in the everyday context Such agents manifest themselves in the form of virtual robots, operating factory processes, humans and embodied mobile robots The advancement of. .. robotics, control, computer science and communications, has made possible the deployment of large teams of mobile robots in real life scenarios These systems have been applied to a wide variety of areas, from manufacturing and warehouse automation, construction and shipping industries, to autonomous robot humanitarian demining, surveillance and urban search -and- rescue A comprehensive overview of issues in multi- robot... decision making within the routine and reactive layers, and further sub-divides these layers into the macro- and micro-decision levels to reflect the degree of coordination required for different processes within a multi- agent team The relationship of these two forms of multi- robot decision making with Ortony’s three layers are also shown in Fig 1.1 The two forms of multi- robot decision making are further... 3.2 3.3 3.4 3.5 Elements concerning an agent, in a Roam-Space with two agents ajk and aj1 k1 The Dandelion Formation and connected graph for agents within and between six Roam-Spaces Agent Clusters and Uncovered Spaces in a Roam-Space Gannt Chart and Convergence of Agents on Roam-Spaces when tasks may end before the expected time or be... background literature and introduction to the main chapters of this thesis 4 1.2 Macro-Level Planning & Inter-Task Coordination Mission/ Task Specifications Robot Team Perceptions and Observations Macro Level Coordination Mechanism Task Dissemination and Allocation Formation of Sub-Teams Robot Sub-Team Robot Sub-Team Micro Level Coordination Action Coordination Micro Level Coordination Action Coordination Individual... organization of tasks into a feasible work plan or schedule The chapter first examines the representation of allocations in the framework of self-organizing virtual agents, with the use of a Dandelion network/graph that defines agent neighborhood relationships Next, algorithms that govern agent interactions and produce clustering behavior are presented The agents self organize to produce a feasible and flexible... planning and inter-task coordination This involves planning the job 3 1.1 Motivation of Research: Multi- Agent Coordination assignments to each robot (or robot sub-team), taking into account their capabilities and the requirements of each job Such mechanisms may be centralized and performed by a single leader robot, or decentralized, in which robots reach a suitable arrangement through a series of observations... considers software agents One major difference is that embodied robots respond and interact directly with other entities in the environment Virtual agents, on the other hand, perform tasks that are mostly informational [4] Furthermore, the constraints faced by software agents are different from those faced by mobile robots, which also influences the type of solution techniques available for each type of agents... Actions Robot Sub-Team … Micro Level Coordination Action Coordination Individual Robot Actions Figure 1.2: Different Levels of Robot Coordination The coordination mechanisms at each level may either be centralized or decentralized 1.2 Macro-Level Planning & Inter-Task Coordination The macro-level of the decision making process deals with the coordination of plans within the agent team such that the overall . well as the outline of this thesis is also presented. 1.1 Motivation of Research: Multi- Agent Coordination Multi- Agent Systems, consisting of huge numbers of interacting agents, linked to- gether. members, and this falls into the realm of micro-level coordination. Such forms of coordination is investigated in the context of representing and cooperative accomplishment of multi- agent formations. Under. CONTROL & COORDINATION OF MULTI- AGENT SYSTEMS CHENG-HENG FUA B. Eng (Hons.), National University of Singapore A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY NUS