RBE PhD Qualifier Presentation
Task Allocation for Dynamic and Static Task Sets in Multi-Agent Systems
Under certain circumstances, static path planning and task allocation algorithms are sufficient for multi-agent systems. However, in the real world in which environments are ever-changing, adaptability is necessary when considering planning algorithms. In order to increase this quality and prepare for real world systems, algorithms must be developed that are able to respond and adapt to a dynamic set of tasks. Several state-of-the-art methods of task allocation in both static and dynamic scenarios exist, including but not limited to space-filling curves, distributed auctions, algorithmic queueing theory, and tree-recoloring. Each one has specific strengths and weaknesses dependent upon the specific system in which they are implemented. This work will discuss these methods, the limitations, and possible future of general task allocation in robotic swarms in a dynamic environment.
Tuesday, December 11, 2018
3:00 p.m. – 4:00 p.m.
85 Prescott Street, RBE Conf Rm 209