/remp

Rearrangement with Multiple Manipulation Primitives

Primary LanguagePython

Toward Optimal Tabletop Rearrangement with Multiple Manipulation Primitives

Baichuan Huang, Xujia Zhang, Jingjin Yu

https://arxiv.org/abs/2310.00167 (ICRA 2024)

3 minutues intro video:

Remp.2.mp4

Installation

  • Docker (amd64)

    git clone https://github.com/arc-l/remp.git
cd remp
docker pull arcldocker/remp:latest
bash run_container.sh
mkdir logs

  • Manually install (Ubuntu)
    Refer to install.sh for a step-by-step installation

Quick Start

  • Case by case run:
    • python environment.py --method mcts --case real_case_6_1 --gui
  • Benchmark:
    • bash benchmark.sh

Test cases

All cases are stored in sim_tests.

Example: Run python environment.py --method mcts --case case_4_3 --gui A PyBullet GUI will pop up, the algorithm will plan the action, and the robot in the simulation will operate.