escher_openrave_cpp

OpenRAVE C++ plugin for Escher Robot to run contact planner and experiment for the paper: "Robust Humanoid Contact Planning with Learned Zero- and One-Step Capturability Prediction" in RA-L 2020.

escher_cpp_motion_planner_interface.py is the main testing script used to run experiments. It will call the corresponding functions in EscherMotionPlanning.cpp through OpenRAVE cpp interface with configurations shown in escher_openrave_cpp_wrapper.py. The contact planner is described in ContactSpacePlanning.cpp.

Dependency

• OpenRAVE 0.9.0
• ROS Melodic/Kinetic/Indigo
• frugally-deep
• Tensorflow 1.40
• comps
• SL

Setup

1. Install ROS, OpenRAVE, Tensorflow (with C++ API), and frugally-deep
2. Create a catkin workspace, and put escher_openrave_cpp, comps and SL under the same catkin workspace.
3. catkin_make to build the code.

Usage

• surface_source: The source of the environment in planning. The user can create new environment by adding new options in update_environment function in environment_handler_2.py.
  capture_test_env_3: Narrow flat corridor environment.
  capture_test_env_4: One-wall rubble environment.
  capture_test_env_5: Oil platform environment.
  load_from_data: Load environment object file stored using pickle from path specified by environment_path parameter.

• environment_path: The folder which contains the stored environment object file.

• start_env_id and end_env_id: The first and last environment object file id loaded in the process.

Example Usage:

python escher_cpp_motion_planner_interface.py surface_source capture_test_env_4 start_env_id 0 end_env_id 0