Skip to content

fluentrobotics/HMPNav

Repository files navigation

MPPI with Human Motion Prediction

This is the control code used in the user study for the paper How Human Motion Prediction Quality Shapes Social Robot Navigation Performance in Constrained Spaces, appearing at the 2026 International Conference on Human Robot Interaction (HRI).

Getting Started

Be in a ROS2 environment (e.g., the fluentrobotics/ros:humble Docker image).

Run poetry install.

Using the Simulator Node

In a dedicated terminal, run rviz2 -d debug.rviz

Make sure the virtual environment is activated: poetry shell

Run python3 -m stretch_mppi.stretch_simulator

Use the 2D Pose Estimate tool at the top of the RViz window to override Stretch's pose (click and drag in the map to set a position and orientation).

Running the MPPI Node

Make sure the virtual environment is activated: poetry shell

Run python3 -m stretch_mppi.prototype_node

Running the HST or CoHAN Prediction Models

To change the prediction model, change the METHOD parameter in controller_config.py. You will need to set up and concurrently run the HST Combo or CoHAN in parallel.

Using the Stretch

The rviz command for the simulator is actually still useful to visualize rollouts, but the state history won't be visualized without running the simulator.

Boot and home the robot.

Launch the stretch driver in navigation mode: ros2 launch stretch_core stretch_driver.launch.py mode:=navigation

Set up motion capture on the dedicated desktop, reconfigure settings on the Stretch if necessary, then run ros2 launch mocap_optitrack_client launch_z_up.py

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages