*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Collisions as Information Sources in Robot Swarms
Committee:
Dr. Egerstedt, Advisor
Dr. Wardi, Chair
Dr. Goldman
Dr. Hutchinson
Abstract:
The objective of the proposed research is to investigate whether inter-robot collisions occurring in robot swarms can be used as an information source which allows the robots to perform useful tasks. We envision a collection of robots equipped with simple tactile collision sensors and demonstrate how these can be used by the robots to perform tasks like localization in a region and decentralized role allocation within the swarm. This work is motivated by the observation that as the size of robots decreases in multi-robot systems, collisions cease to be catastrophic events that need to be avoided at all costs. This implies that less conservative, coordinated control strategies can be employed, where collisions are not only tolerated, but can potentially be harnessed as an information source. Moreover, many animal species have been shown to use physical contact within a swarm as a sensory mechanism to perform tasks. In this proposal, we follow this line of inquiry by employing collisions as a sensing modality that provides information about the robots' surroundings. To this end, we develop theoretical models to analyze the nature and frequency of collisions occurring between robots by using ideas from stochastic geometry and molecular physics. In the first application, we investigate whether the collisions are sufficiently rich sources of information to allow the robots to localize themselves (in particular environments). Following this, we develop a role allocation strategy which allows robots in a swarm to distribute themselves among different roles using simple collision-based interactions. The resulting algorithm is completely decentralized and is robust to robot failures within the swarm. The developed algorithms are deployed on teams of real-world robots. Inspired by these developments, we propose three directions of future work where we would like to explore connections between collisions occurring among swarming robots, in insect colonies, and among molecules in a physical system.
