Swarm Robotics: Second SAB 2006 International Workshop, Rome, Italy, September 30-October 1, 2006 Revised Selected Papers (Lecture Notes in Computer Science (4433))

Swarm robotics is the study of how large numbers of relatively simple physically embodied agentscanbe designedsuchthat a desiredcollectivebehavioremerges from the local interactions among agents and between the agents and the en- ronment. Swarm robotics has emerged as a novel approach to the coordination of large numbers of robots and is inspired from observation of social insects – ants, termites, wasps and bees – which stand as fascinating examples of how a large number of simple individuals can interact to create collectively intelligent systems. Socialinsects areknownto coordinatetheir actionsto accomplishtasks that are far beyond the capabilities of a single individual: termites build large and complex mounds, army ants organize impressive foraging raids, ants can collectively carry large prey. Such coordination capabilities are still well beyond the reach of current multi-robot systems. Researchon swarmroboticshasseen a signi?cantincreasein the last 5 years. A number of successful swarm robotic systems have now been demonstrated in the laboratory and the study of the design, modelling, implementation and analysis of swarm robotic systems has become a hot topic of research. This workshop was organized within SAB 2006, as a sequel to the successful ?rst swarmroboticsworkshopin2004,withtheaimofreviewingandupdatingrecent advances on the topic. We received 21 full papers (20 research + 1 review) and accepted 14 (13 research + 1 review). Overall, we, as organizers, were pleased with the number of submissions, and a number of our reviewers explicitly commented on the generally high quality of the papers. Title page Preface Organization Table of Contents A Navigation Algorithm for Swarm Robotics Inspired by Slime Mold Aggregation Motivation The Simulation Platform and the Swarm Scenario The ‘Slime Mold’ Strategy Detailed Description of the Strategy and of the Simulator Results Evolving Optimal Parameters Evaluation of Critical Parameters Alternative ‘Good’ Parameter Settings Way Finding and Trail Formation The ‘Gradient Version’ of the Slime Mold Algorithm Discussion Strategies for Energy Optimisation in a Swarm of Foraging Robots Introduction Adaptation Mechanism Experimental Set-Up Experimental Results and Analysis Conclusion A Macroscopic Model for Self-organized Aggregation in Swarm Robotic Systems Introduction Related Studies Aggregation Modeling Aggregation Problem Aggregation Behavior A Macroscopic Model for Aggregation Probabilities for Shrinking Probabilities for Growth Probabilities to Remain Same Macroscopic Model Experimental Results Conclusion An Analytical and Spatial Model of Foraging in a Swarm of Robots Introduction Simulation Analytical Model Results Conclusion and Outlook Algorithms for the Analysis and Synthesis of a Bio-inspired Swarm Robotic System Introduction Methodology Macro-continuous Model Reachability Analysis Algorithm Application to the House-Hunting Model Simulation Algorithms Application to the House-Hunting Model Conclusion Coordination and Control of Multi-agent Dynamic Systems: Models and Approaches Introduction Multi-agent Dynamic Systems, Formations, and Swarms Formation Control Graphs and Underlying Graphs Rigidity and Persistence Neighborhood Mathematical Models for Agent Dynamics Higher-Level (Single Integrator) Model Point Mass (Double Integrator) Model Fully Actuated Model with Uncertainty Non-holonomic Unicycle Model Dubins' Vehicle Model Self-propelled Particle Model Swarm Coordination and Control Problems Aggregation and Social Foraging Flocking and Rendezvous Formation Stabilization and Acquisition Formation Maintenance and Cohesive Motion Control Formation Reconfiguration and Switching Distributed Agreement Problems Cooperative Control Approaches to Modeling and Coordination and Control of Swarms Potential Function Based Approaches Sliding Mode Control Feedback Linearization Lyapunov Analysis and Other Nonlinear Control Techniques Behavior Based and Evolutionary Approaches Artificial Physics Asynchronous Swarm Models Probabilistic Approaches Stability, Performance, and Robustness Concluding Remarks Communication in a Swarm of Miniature Robots: The e-Puck as an Educational Tool for Swarm Robotics Introduction Tools and Methods for Swarm Robotics The e-Puck: An Educational Robot The Webots^{TM}Simulation Environment Correspondence Between Reality and Simulation Communication and Swarm Intelligence A Radio Communication Module for the e-Puck Hardware Design and Structure of the Module Software Control of the Radio Board Measurement of Physical Characteristics Case Study: Collective Decision Experimental Setup Results Related Hybrid Network Example: Isolated Collective Decision Conclusion UltraSwarm: A Further Step Towards a Flock of Miniature Helicopters Introduction The Idea The Helicopter Platform Electronics and Sensors State Estimation An Automated Design Method Model Identification Controller Design Future Work Concluding Remarks Where Are You? Goal of Our Work Localization Measuring Distance Channeling Acoustic Energy into a Plane Related Work Trilateration Method I Trilateration Method II Trilateration Method II + Communication Trilateration Implementation Trilateration Hardware Synchronization Protocol The Maxelbot Experiments and Demonstrations Accuracy Experiment Linear Formations Box/Baby Pulling Physicomimetics Formations for Chemical Plume Tracing Summary Collective Perception in a Robot Swarm Introduction The Scenario The Hop-Count Strategy The ‘Trophallaxis-inspired’ Strategy Results Scaling the Sizes of the Target Areas The Importance of the Swarm Density The Role of the Aggregation Threshold $(th_{agg})$ The Role of the Negative Feedback $(r_c)$ Discussion References Distributed Task Selection in Multi-agent Based Swarms Using Heuristic Strategies Introduction Multi-agent Swarming Task Selection in Swarming Distributed Task Selection Model Heuristic-Based Task Selection Strategies Simulation Results Experimental Setup Experimental Results Related Work Conclusion and Future Work Evolution of Signalling in a Group of Robots Controlled by Dynamic Neural Networks Introduction Description of the Task The Simulation Model The Controller and the Evolutionary Algorithm The Fitness Function Results Porting on Real Robots Conclusions Collective Specialization for Evolutionary Design of a Multi-robot System Introduction Research Goal Specialization First Hypothesis Second Hypothesis CONE: Collective Neuro-Evolution Online Versus Offline Adaptation in the CONE Method Conventional Neuro-Evolution Genotypes Recombination of Genotypes: Crossover and Mutation Fitness Calculation Phenotypes: Constructing Controllers from Neurons Dynamic Topologies Collective Survey Task Environment Red Rocks (Features of Interest) Red Rock Distribution Agents Morphology: Sensors and Actuators Controllers: Adaptive Topology Neural Network Heuristic Methods Specialized and Non-specialized Group Types Red Rock Discovery Algorithm Experiments and Results Heuristic Method Comparison: Specialized Versus Non-specialized Groups Neuro-Evolution Method Comparison Analysis and Discussion Conclusions Scalability in Evolved Neurocontrollers That Guide a Swarm of Robots in a Navigation Task Introduction The Simulated Agents The Controller and the Evolutionary Algorithm The Fitness Function Results Scalability Results Analysis of Controller Dynamics for a Single Agent Analysis of Controller Dynamics for the Entire Team Conclusions Author Index

This book constitutes the thoroughly refereed post-proceedings of the 2nd SAB 2006 International Workshop on Swarm Robotics held in Rome, Italy in September/October 2006 as a satellite event of SAB 2006, the 9th Conference on Simulation of Adaptive Behavior.

The 14 revised full papers presented were carefully reviewed and selected from 21 submissions. The papers are organized in topical sections on algorithms, modelling and analysis, hardware, and evolutionary approaches.

This Book Constitutes The Thoroughly Refereed Post-proceedings Of The 2nd Sab 2006 International Workshop On Swarm Robotics Held In Rome, Italy In September/october 2006 As A Satellite Event Of Sab 2006, The 9th Conference On Simulation Of Adaptive Behavior. The 14 Revised Full Papers Are Organized In Topical Sections On Algorithms, Modeling And Analysis, Hardware, And Evolutionary Approaches.