Research & Papers

Competition - League of Robot Runners 2026: Multi-robot coordination under uncertainty [N]

Can your RL algorithm coordinate thousands of robots under real-time uncertainty?

Deep Dive

The League of Robot Runners (LoRR) 2026, co-located with the AAMAS 2026 conference, is a research competition focused on large-scale multi-robot coordination under uncertainty. Participants must control hundreds to thousands of robots that work together to complete tasks, move efficiently across diverse maps, and operate continuously in real-time. The problem is made especially challenging by nested combinatorial decision-making (task assignment + path planning) and movement uncertainty—robot actions have a probability of being delayed. The organizers specifically highlight that ML and RL methods are well-suited for these policy-based, stochastic environments, making this a prime opportunity to benchmark novel algorithms against symbolic and OR approaches.

Competitors can enter one or more of three distinct tracks: Task Scheduling, Execution, and Combined, with cash prizes for top submissions. A starter kit in C++ and Python, example instances, validators, and a visualizer are provided. Submissions are evaluated automatically with live leaderboard feedback throughout the main round (April 16 to July 22, 2026). The AAMAS prize deadline is May 22, 2026. All techniques—search/planning, RL/ML, operations research, mathematical programming, robust optimization, and hybrids—are welcome. This competition directly addresses real-world challenges in logistics, manufacturing, and computer games, where scalable, robust multi-robot coordination is critical.

Key Points
  • Competition involves coordinating hundreds to thousands of robots in real-time under movement uncertainty (actions have delay probabilities).
  • Three tracks: Task Scheduling, Execution, and Combined – each with cash prizes and auto-evaluated live leaderboard.
  • Timeline: main round April 16 to July 22, 2026; AAMAS prize deadline May 22, 2026; starter kit (C++/Python) and visualizer provided.

Why It Matters

Multi-robot coordination drives logistics, manufacturing, and games; this competition pushes ML/RL to solve real-world uncertainty at scale.