Welcome to SMART

SMART is an open-source testbed for executing and evaluating Multi-Agent Path Finding (MAPF) plans in a realistic simulator. The public codebase combines ARGoS 3 for simulation, an Action Dependency Graph (ADG) based execution monitor, and helper scripts for turning benchmark maps and scenarios into runnable experiments.

This documentation site tracks the current public SMART sources:

Video Demo

Watch the SMART demonstration video on YouTube: https://www.youtube.com/watch?v=TX-oGSgM8VQ

How to Get It

The public repository is distributed as source code:

How to Use It

SMART provides a few practical entry points:

What the Public Repo Includes

The current public SMART repository contains:

  • run_sim.py as the main helper script

  • ArgosConfig/ for generating ARGoS configuration files

  • server/ for the ADG execution-monitoring server

  • client/ for ARGoS controllers and executors

  • Example maps, scenarios, and path files in the repository root

Research Paper

The technical background and evaluation are described in the SMART paper: Advancing MAPF towards the Real World: A Scalable Multi-Agent Realistic Testbed (SMART).

Please cite it as:

@article{yan2025smart,
  title={Advancing MAPF towards the Real World: A Scalable Multi-Agent Realistic Testbed (SMART)},
  author={Yan, Jingtian and Li, Zhifei and Kang, William and Zheng, Kevin and Zhang, Yulun and Chen, Zhe and Zhang, Yue and Harabor, Daniel and Smith, Stephen F and Li, Jiaoyang},
  journal={arXiv preprint arXiv:2503.04798},
  year={2025}
}

Participate

FAQ

If you run into problems, check FAQ - Frequently Asked Questions.

License

SMART is released under the MIT License.

Copyright (c) 2025 Carnegie Mellon University

Getting Started