Multi-agent systems toolbox

Toolbox containing several MAS environments, allowing to compare different algorithms

Description:

Toolbox for testing and evaluating multi-agent control systems on two warehouse-inspired environments: Multi-agent Pick Up And Delivery (MAPD), Multi-agent Pathfinding (MAPF). It can be used to benchmark the performance of several multi-agent planning and learning methods.

More Information:

Originates from the Collaborative AI research challenge (FAIR 2019–2023).
Responsible Research Lead: Ann Nowé
Authors: Marjon Blondeel
Programming Language: C#/Python
Link to the website: https://ai.vub.ac.be/toolbox-multi-agent-systems/
Link to the repository: MAPD (https://gitlab.ai.vub.ac.be/multi-agent-benchmarking/pick-up-and-delivery), MAPF (https://gitlab.ai.vub.ac.be/multi-agent-benchmarking/path-finding)