Dynamic Routing and Coordination in Multi-Agent Networks
Abstract
Supported by this project, we designed innovative routing, planning and coordination strategies for robotic networks and studied their application to Army and DoD scenarios. The key technological challenge is the decision of who does what, when or, equivalently, how are tasks partitioned among robots, in what order are they to be performed, and along which deterministic routes or according to which stochastic rules do individual robots move. The fundamental novelties and our recent breakthroughs supported by this project are manifold: (1) the application of queueing theory and combinatorial techniques to network of autonomous robots leads to a wide range of new relevant problems, (2) novel coordination schemes promise to successfully achieve various optimization objectives relying only upon asynchronous and asymmetric communication, (3) the increasingly weaker assumptions imposed on routing and coordination algorithms are rendering them practical and widely applicable. This project addressed multi-dimensional problems of relevance in Engineering and Computer Science by unifying fundamental concepts from multiple domains (robotics, autonomy, combinatorics, and network science). Our work aimed to bridge multiple scientific disciplines, including control theory and theoretical computer science and their applications to multi-agent systems, robotics and sensor networks.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 10, 2016
- Accession Number
- AD1010873
Entities
People
- Francesco Bullo
Organizations
- University of California, Santa Barbara