COAgents: Multi-Agent Framework to Learn and Navigate Routing Problems Search Space
Researchers have developed COAgents, a novel multi-agent framework designed to tackle complex Vehicle Routing Problems (VRPs). This framework models the search for optimal solutions as a graph, using specialized agents to guide exploration and diversification. COAgents demonstrates strong performance on CVRP benchmarks and achieves state-of-the-art results among learning-based methods for the more challenging VRPTW instances. AI
IMPACT Introduces a new learning-based approach that sets state-of-the-art performance on challenging routing problems.