Ph.D. Research Proposal Exam: Senthil Hariharan Arul
Wednesday, November 30, 2022
12:00 p.m.-2:00 p.m.
301 405 0680
ANNOUNCEMENT: Ph.D. Research Proposal Exam
Name: Senthil Hariharan Arul
Prof. Dinesh Manocha (Chair)
Prof. Kaiqing Zhang
Prof. Pratap Tokekar
Date/time: Nov 30, 2022 (12-2pm)
Location: AVW 2328
Title: Decentralized Multi-Agent Navigation in Complex Scenarios
Multi-robot systems (MRS) have received significant attention owing to their wide-ranging applications, such as warehouse automation, search and rescue, surveillance, and last-mile delivery. Navigation is an integral part of these applications, which involve the robots moving safely, and efficiently and cooperating with other robots in the MRS—our research is centered around developing decentralized methods for generating safe and efficient multi-robot navigation. Our research on multi-agent navigation is classified into three parts; In the first part, we present approaches for quadrotor or UAV navigation in multi-agent, complex scenarios and subsequently extend it to handle sensing dynamics constraints and uncertainty. In the second part, we consider dense multi-agent environments, where we propose two methods for navigating ground robots in congested scenes. In the third part, we describe new methods that exploit reinforcement learning (RL) to improve the performance of decentralized navigation. In particular, we present our current approach to multi-agent navigation using inter-agent communication and discuss our ongoing work on RL-based navigation.