MAPF Visualizer
About The Project
This project provides a visualization tool for Multi-Agent Path Finding (MAPF) algorithms.
Many single-agent pathfinding visualizers showcase classic algorithms such as A* and Dijkstra. However, the field of multi-agent path finding is relatively new (CBS, an important MAPF algorithm, was proposed in 2012) and thus didn’t gain as much public attention.
This website aims at helping people better understand MAPF by offering a real-time visualization tool. Usually running a MAPF solver involves the following steps:
- compile the C++ code into executables
- put the map and instance into two separate files with contents formatted in terms of certain rules
- run the executable with a complicated command
This website offers a much more intuitive experience. Users will be able to:
- select a particular algorithm they are interested in
- design their own map by dragging their mouse to add walls
- add agents by entering their start and goal locations
- press the
planbutton and get the animated planning result instantly
Built With
This project is bootstrapped with the following frameworks and libraries:
Getting Started
Open MAPF Visualizer in one of the following browsers for optimal support:
- Chrome v98 and later
- Firefox v94 or later
- Edge v98 or later
- Safari v15.4 or later
Contributing
If you have a MAPF-related algorithm that might fit this project, feel free to email me—I’m glad to collaborate on integrating it.
More to implement
- Add pages for more detailed information about MAPF algorithms and corresponding papers.
- Include more MAPF algorithms (CBS-based, SAT-based, etc.) for users to run their instances with.
- Include some other MAPF variants, such as k-robust and lifelong MAPF.
- Auto-load some benchmark examples.
- …
Github repo
If you find it useful, consider starring the repo on GitHub.
License
Distributed under the MIT License.