Pushing Through Clutter With Movability Awareness of Blocking Obstacles

• Under review for the International Conference on Robotics and Automation (ICRA) 2025

Abstract

Navigation Among Movable Obstacles (NAMO) poses a challenge for traditional path-planning methods when obstacles block the path, requiring push actions to reach the goal. We propose a framework that enables movability-aware planning to overcome this challenge without relying on explicit obstacle placement. Our framework integrates a global Semantic Visibility Graph and a local Model Predictive Path Integral (SVG-MPPI) approach to efficiently sample rollouts, taking into account the continuous range of obstacle movability. A physics engine is adopted to simulate the interaction result of the rollouts with the environment, and generate trajectories that minimize contact force. In qualitative and quantitative experiments, SVG-MPPI outperforms the existing paradigm that uses only binary movability for planning, achieving higher success rates with reduced cumulative contact forces.

Overview

An overview of the proposed SVG-MPPI architecture where the SVG provides a weighted graph with efficient node placement around movable obstacles along which a lowest-effort path can be found. The generated set of waypoints guides the MPPI control strategy to efficiently sample rollouts around movable obstacles. If during interaction an obstacle is considered non-movable, the movability estimation gets updated and the path is replanned. Snapshots of a real-world example are shown on the left where the red star indicates the goal location and the masses of the obstacles are (A): 25 kg, (B): 20 kg, (C): 5 kg.

Example

Installation

To get the project up and running, follow the steps below:

1. Set Interpreter to Poetry Environment:

   • Ensure that your Python interpreter is correctly set to the Poetry virtual environment. If you haven't set up Poetry, install it by following Poetry's documentation.
   • Run the following command to activate the Poetry environment for this project:

     poetry install

   • Confirm that Poetry is using the correct environment by checking the executable path:

     poetry env info --executable

2. Download and Install Isaac Gym:

   • The project relies on Isaac Gym for its physics simulation capabilities. You need to download Isaac Gym from NVIDIA's website.
   • Once downloaded, place the isaacgym folder inside the src directory of the project:

     project-root/
     ├── src/
     │   └── isaacgym/
     

   • Follow the installation instructions provided by NVIDIA to ensure Isaac Gym is correctly set up in your environment.

3. Update Git Submodules:

   • The project might include external dependencies managed as Git submodules. Make sure all submodules are updated by running the following commands:

     git submodule init
     git submodule update --recursive --remote

   • If new submodules have been added, this ensures they are properly fetched and integrated into the project.

4. Build the Project:

   • Once the environment and dependencies are ready, build the project. Navigate to the project’s root directory and run:

     poetry run build

   • This command will ensure that all necessary components are compiled and ready for use.

5. Set Up ROS:

   • If your project interacts with ROS (Robot Operating System), make sure to source the correct ROS environment. Run the following command to source the environment based on your ROS distribution (e.g., melodic, noetic):

     source /opt/ros/<ros_distro>/setup.bash

   • Make sure your ROS workspace is also built and sourced before running the project:

     catkin_make
     source devel/setup.bash

6. Verify Installation:

   • Test the installation by running one of the provided examples or demos to ensure all components are functioning as expected. If any issues arise, check the logs or error messages for more information.

Troubleshooting

• To check the environment executable path, use:

  poetry env info --executable

• If running out of space, clear poetry cache:

    poetry cache clear --all .