Autonomous Intelligence and Robotics (AIRob) Lab

Autonomous Intelligence and Robotics (AIROB) Lab is interested in topics on automated planning, multi-agent/robot systems, spatio-temporal and constraint reasoning, and applications of probabilistic methods and other topics related to graphs, combinatorial optimization, and algorithms.

Areas of Research

Teams of agents often need to continuously assign tasks among themselves and plan collision-free paths to task locations. For example, in collaboration with researchers from NASA's Ames Research Center, the AIROB Lab envisions a future where autonomous aircraft towing vehicles transport aircraft from runways to gates (and vice versa), reducing pollution, energy consumption, congestion, and human workload. With support from Amazon Robotics and Alibaba, the lab addresses computational challenges in scenarios where hundreds of robots autonomously navigate warehouses, moving inventory pods from storage locations to inventory stations (and vice versa).

The coordination of these agents poses NP-hard problems, requiring real-time task assignment and high-quality, collision-free pathfinding. The AIROB Lab studies various versions of task-assignment (TA) and multi-agent pathfinding (MAPF) problems, exploring their complexities, algorithms for solving them, and their applications. The lab integrates tools and techniques from artificial intelligence, robotics, operations research, and theoretical computer science to establish an algorithmic foundation for making fast, optimal decisions to coordinate long-term task and path-planning operations for real-world multi-agent systems on a large scale, involving hundreds of agents and thousands of tasks.