pibt2

The code repository of the paper "Priority Inheritance with Backtracking for Iterative Multi-agent Path Finding" (AIJ-22; the journal version).

Updated

• 2023-10-02: The visualizer for Ubuntu is available here.
• 2023-07-26: A minimal Python implementation is now available.

Description

• There is an old implementation for the conference paper (IJCAI-19); this repo is much faster and cleaner.
• The implementation includes both Multi-Agent Path Finding (MAPF) and Multi-Agent Pickup and Delivery (MAPD).
• It is written in C++(17) with CMake (≥v3.16) build.
• The repository uses Google Test and the original library for 2D pathfinding as git submodules.
• The visualizer uses openFrameworks and works only on macOS (as git submodule).
• The accompanied solvers are HCA* [1], Push and Swap [2], TP [3], and PIBT(+).

platform	status (public)	status (dev)
macos-10.15
ubuntu-latest

Please cite the following paper if you use the code in your published research:

# AIJ-22

@article{okumura2022priority,
  title = {Priority Inheritance with Backtracking for Iterative Multi-agent Path Finding},
  journal = {Artificial Intelligence},
  pages = {103752},
  year = {2022},
  issn = {0004-3702},
  doi = {https://doi.org/10.1016/j.artint.2022.103752},
  author = {Keisuke Okumura and Manao Machida and Xavier Défago and Yasumasa Tamura},
}

# IJCAI-19

@inproceedings{okumura2019priority,
  title={Priority Inheritance with Backtracking for Iterative Multi-agent Path Finding},
  author={Okumura, Keisuke and Machida, Manao and D{\'e}fago, Xavier and Tamura, Yasumasa},
  booktitle={Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, {IJCAI-19}},
  publisher={International Joint Conferences on Artificial Intelligence Organization},
  pages={535--542},
  year={2019},
  month={7},
  doi={10.24963/ijcai.2019/76},
  url={https://doi.org/10.24963/ijcai.2019/76}
}

Demo

Building

git clone --recursive https://github.com/Kei18/pibt2.git
cd pibt2
mkdir build && cd build
cmake ..
make

Docker

You can also use the docker environment instead of the native one, based on Ubuntu18.04.

docker compose up -d
docker compose exec dev bash

Usage

MAPF

PIBT

./mapf -i ../instances/mapf/sample.txt -s PIBT -o result.txt -v

You can find details and explanations for all parameters with:

./mapf --help

Please see instances/mapf/sample.txt for parameters of instances, e.g., filed, number of agents, time limit, etc.

Output File

This is an example output of ../instances/mapf/sample.txt. Note that (x, y) denotes location. (0, 0) is the left-top point. (x, 0) is the location at x-th column and 1st row.

instance=../instances/mapf/sample.txt
agents=100
map_file=arena.map
solver=PIBT
solved=1
soc=3738
lb_soc=3243
makespan=68
lb_makespan=68
comp_time=11
starts=(32,21),(40,4),(20,22),(26,18), [...]
goals=(10,16),(30,21),(11,42),(44,6), [...]
solution=
0:(32,21),(40,4),(20,22),(26,18), [...]
1:(31,21),(40,5),(20,23),(27,18), [...]
[...]

MAPD

PIBT

./mapd -i ../instances/mapd/sample.txt -s PIBT -o result.txt -v

You can find details and explanations for all parameters with:

./mapd --help

Please see instances/mapd/sample.txt for parameters of instances.

When you specify pickup and delivery locations (and non-task endpoints), put a special file in map/. An example is map/warehouse.map.pd. The rule is the following:

• [@T]: obstacles
• [psa]: pickup locations
• [dsa] delivery locations
• [ea] endpoints

Output File

This is an example output of ../instances/mapd/sample.txt.

• task: {task-id}:{loc-pickup}->{loc-delivery},appear={timestep},finished={timestep}
• solution: {timestep}:(loc-current)->(loc-current-target):{assigned-task-id or -1}

instance=../instances/mapd/sample.txt
agents=50
map_file=warehouse.map
solver=PIBT
solved=1
service_time=25.33
makespan=535
comp_time=437
preprocessing_comp_time0
starts=(33,17),(32,19), [...]
task=
0:129->113,appear=0,finished=23
4:342->256,appear=4,finished=24
[...]
solution=
0:(33,17)->(24,3):-1,(32,19)->(24,3):-1, [...]
1:(32,17)->(29,2):-1,(31,19)->(29,2):-1, [...]
[...]

Visualizer

Building

It takes around 10 minutes.

macOS 10.x

bash ./visualizer/scripts/build_macos.sh

Note: The script of openFrameworks seems to contain bugs. Check this issue. I fixed this in my script :D

macOS 11.x

git submodule update --remote
bash ./third_party/openFrameworks/scripts/osx/download_libs.sh
cd visualizer
make build
cd ..
chmod +x ./visualize.sh

Usage

cd build
../visualize.sh result.txt

You can manipulate it via your keyboard. See printed info.

Experimental Environment

The experimental scripts are written in Python3.7.

Exp	used version	scripts
MAPF		exp_scripts/mapf.py
MAPD		exp_scripts/mapd.py

Notes

• Maps in maps/ are from MAPF benchmarks. When you add a new map, please place it in the maps/ directory.
• The font in visualizer/bin/data is from Google Fonts.
• Other baseline solvers are obtained from: CBS, EECBS, and BCP

Licence

This software is released under the MIT License, see LICENSE.txt.

Author

Keisuke Okumura is a Ph.D. student at the Tokyo Institute of Technology, interested in controlling multiple moving agents.

Reference

1. Silver, D. (2005). Cooperative pathfinding. Proc. AAAI Conf. on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-05)
2. Luna, R., & Bekris, K. E. (2011). Push and swap: Fast cooperative path-finding with completeness guarantees. Proc. Int. Joint Conf. on Artificial Intelligence (IJCAI)
3. Ma, H., Li, J., Kumar, T. K., & Koenig, S. (2017). Lifelong multi-agent path finding for online pickup and delivery tasks. Proc. Int. Conf. on Autonomous Agents and Multiagent Systems (AAMAS)