Robotics Seminar

  • Gates Hillman Centers
  • Mauldin Auditorium 1305
  • Professor
  • Computer Science Department
  • University of Southern California

Progress on Multi-Robot Path Finding

Teams of robots often have to assign target locations among themselves and then plan collision-free paths to their target locations. Examples include autonomous aircraft towing vehicles and automated warehouse systems. For example, in the near future, autonomous aircraft towing vehicles might tow aircraft all the way from the runways to their gates (and vice versa), thereby reducing pollution, energy consumption, congestion and human workload. Today, hundreds of robots already navigate autonomously in Amazon fulfillment centers to move inventory pods all the way from their storage locations to the packing stations. Path planning for these robots can be NP-hard, yet one must find high-quality collision-free paths for them in real-time. The shorter these paths are, the fewer robots are needed and the cheaper it is to open new fulfillment centers. In this talk, I describe several variants of the multi-robot path-planning problem, their complexities and algorithms for solving them. I also present a hierarchical planning architecture that combines ideas from artificial intelligence and robotics. It makes use of a simple temporal network to post-process the output of a multi-robot path-finding algorithm in polynomial time to create a plan-execution schedule that take the maximum translational and rotational velocities of non-holonomic robots into account, provides a guaranteed safety distance between them, and exploits slack to absorb imperfect plan executions and avoid time-intensive re-planning in many cases.

This research is joint research with N. Ayanian, T. Cai, L. Cohen, W. Hoenig, S. Kumar, H. Ma, G. Sharon, C. Tovey, T. Uras, H. Xu, S. Young, D. Zhang, and other colleagues and students.

Sven Koenig is a professor in computer science at the University of Southern California. Most of his research centers around techniques for decision making (planning and learning) that enable single situated agents (such as robots or decision-support systems) and teams of agents to act intelligently in their environments and exhibit goal-directed behavior in real-time, even if they have only incomplete knowledge of their environment, imperfect abilities to manipulate it, limited or noisy perception or insufficient reasoning speed. He co-founded both Robotics: Science and Systems and the International Symposium on Combinatorial Search (SoCS) and was conference or program co-chair of SARA 2002, ICAPS 2004, AAMAS 2005, SoCS 2009, AAAI 2015, and EAAI 2016 and 2017. He is chair of the ACM Special Interest Group on Artificial Intelligence, an editor of AI Magazine and CACM, and an associate editor of AIJ, JAAMAS, and ACS. He was a councilor of AAAI, a member of the advisory board of JAIR, an associate editor of Computational Intelligence, and a member of the steering committees of ICAPS, SoCS, and SARA.

For More Information, Please Contact: