Skip to content

A task space motion planner with global and local obstacle avoidance for robotic manipulators.

License

Notifications You must be signed in to change notification settings

Tom-Forsyth/LCPMotionPlanner

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LCPMotionPlanner

Summary

LCPMotionPlanner is a task space motion planner with complementarity-based obstacle avoidance for spatial manipulators.

The planning framework consists of a global and local planner. The global planner is an RRT planner that will sample poses in the task space. The local planner, also a task space planner, is screw linear interpolation (ScLERP) based, and has obstacle avoidance formulated as a linear complementarity problem (LCP).

Movement of the manipulator to avoid obstacles is projected onto the null space of the jacobian, which ensures end-effector constraints will always be satisfied, even while actively avoiding obstacles.

The benefits of this approach over a traditional joint space RRT planner are computational efficiency, ease of end-effector constraint enforcement, and natural obstacle avoidance motions.

Supported Platforms

Currently, only Windows and Linux are supported. For Windows, builds have been tested on Windows 11 with MSVC 17. For Linux, builds have been tested on Ubuntu 20.04 with GCC 9.4.0.

Documentation

To generate documentation, run doxygen using the supplied Doxyfile, and open index.html.

cd docs/doxygen
doxygen Doxyfile
cd html && firefox index.html

Dependencies

LCPMotionPlanner has three dependencies - Eigen, PhysX, and Boost. Eigen and PhysX are included as submodules of this repository and handled by CMake, so no extra effort is required to get started. Due to the large size of the Boost repository, a local installation will be needed. Please download the header files, extract the files, and set the location in the root CMakeLists.txt.

# Boost.
set(BOOST_ROOT "C:/src/boost_1_80_0") # Your location here.
set(BOOST_INCLUDEDIR "${BOOST_ROOT}/boost")
find_package(Boost REQUIRED)

If you are on Linux, you may need some extra packages for PhysX. Please install the packages below.

sudo apt-get install build-essential libxxf86vm-dev libgl1-mesa-dev libglu1-mesa-dev gcc-multilib g++-multilib freeglut3-dev lib32z1

Build Configurations

PhysX ships with four build configurations. Prefer to use the debug or release mode as they are most stable with LCPMotionPlanner.

  • "debug": Full debug build of PhysX + LCPMotionPlanner with checks and visualization.
  • "checked": Release build of PhysX with checks and visualization + release build of LCPMotionPlanner with debugging.
  • "profile": Release build of PhysX without chceks but still with visualization + release build of LCPMotionPlanner with debugging.
  • "release": Full release build of PhysX + LCPMotionPlanner with no visualization or debugging.

Build Instructions

Clone the repository and submodules.

git clone [email protected]:Tom-Forsyth/LCPMotionPlanner.git
cd LCPMotionPlanner
git submodule update --recursive --init

Run CMake with the desired build configuration to generate a makefile (Linux) or Visual Studio solution (Windows).

cmake -B build -DCMAKE_BUILD_TYPE=release

If you are on Linux, build the project using the make file, and run the simulation.

cd build
make -j20
./LCPMotionPlanner

If you are on Windows, open "LCPMotionPlanner.sln" inside of the build directory. Press F7 to build the solution, and run the simulation.

Debug Visualization

Currently, debug visualization is only supported through PhysX Visual Debugger, which is only available for Windows. With PhysX Visual Debugger open, run the simulation, and the primatives of the scene will appear in the PVD player. Release builds do not support debug visualization.

High Fidelity Visualization

Realistic visualation is available through UnrealSim, where we are using the API of this library to generate collision free plans for the Franka Panda manipulator. Objects can be placed into the UnrealSim simulation scene, and the objects will then be loaded in as primatives into this library to generate motion plans.

Example Simulation

An example simulation for the Franka Panda is provided.

NVIDIA PhysX SDK 4.1 Redistribution Notice

Copyright (c) 2021 NVIDIA Corporation. All rights reserved.

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

  • Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
  • Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
  • Neither the name of NVIDIA CORPORATION nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.