Prev-Talk | Next-Talk | All-Talks | Talks-Sorted | MOOS-DAWG'24

Talk-27: Multi-agent decision-making with MOOS-IvP

Tyler Paine, Michael Benjamin, Michael Defilippo, Jack Hamel, Raymond Turrisi, MIT Dept of Mechanical Engineering, Marine Autonomy Lab

Multi-agent decision-making with MOOS-IvP is an active area of research at the MIT Marine Autonomy Lab.  In this demo we will show a compilation of methods for group decision-making that include Voronoi-based area coverage, auctioning-based task allocation, and group option selection via consensus and dissensus.  The demonstration scenario is inspired by a high value unit (HVU) protection mission.  The high value asset will be the lab’s 16ft WAM-V by OPT and we expect to use 8-10 Heron USVs made by Clearpath Robotics as the protection agents.  At least one intruder will be detected during the mission and the group decides who will intercept.


Categories:

  • Autonomy