Communications of the ACM - Robots: intelligence, versatility, adaptivity
Multimode locomotion via SuperBot reconfigurable robots
Autonomous Robots
Reaching a Consensus in a Dynamically Changing Environment: A Graphical Approach
SIAM Journal on Control and Optimization
Self-adapting modular robotics: a generalized distributed consensus framework
ICRA'09 Proceedings of the 2009 IEEE international conference on Robotics and Automation
Engineering self-adaptive modular robotics: a bio-inspired approach
IROS'09 Proceedings of the 2009 IEEE/RSJ international conference on Intelligent robots and systems
Collective decision-making in multi-agent systems by implicit leadership
Proceedings of the 9th International Conference on Autonomous Agents and Multiagent Systems: volume 3 - Volume 3
Distributed multiagent learning with a broadcast adaptive subgradient method
Proceedings of the 9th International Conference on Autonomous Agents and Multiagent Systems: volume 1 - Volume 1
A Self-adaptive Framework for Modular Robots in a Dynamic Environment: Theory and Applications
International Journal of Robotics Research
Consensus acceleration in multiagent systems with the Chebyshev semi-iterative method
The 10th International Conference on Autonomous Agents and Multiagent Systems - Volume 1
| Hi-index | 0.00 |
This paper presents a theoretical study of decentralized control for sensing-based shape formation on modular multi-robot systems, where the desired shape is specified in terms of local sensor constraints between neighboring robot agents. We show that this problem can be formulated more generally as distributed constraint-maintenance on a networked multi-agent system. It is strongly related to a class of multi-agent algorithms called distributed consensus, which includes several bio-inspired algorithms such as flocking and firefly synchronization. By exploiting this connection, we can theoretically analyze several important aspects of the decentralized shape formation algorithm and generalize it to more complex multi-agent scenarios. We show that the convergence time depends on (a) the number of robot agents and agent connection topology, (b) the complexity of the user-specified goal, and (c) the initial state of the robots. Using these results, we can provide precise statements on how the approach scales, and how quickly the system can adapt to perturbations. These results provide a deeper understanding of the contrast between centralized and decentralized multi-agent algorithms.