Gradient-Informed Path Smoothing for Wheeled Mobile Robots (GRIPS)

C++ implementation of a post-smoothing approach that improves the quality of paths generated by sampling-based planners.

Planning smooth trajectories is important for the safe, efficient and comfortable operation of mobile robots, such as wheeled robots moving in crowded environments or cars moving at high speed. Asymptotically optimal sampling-based motion planners can be used to generate such trajectories eventually. However, to achieve the necessary efficiency for the real-time operation of robots, one often uses their initial feasible trajectories or the trajectories of non-optimal planners instead, typically after a post-smoothing step. We propose a gradient-informed post-smoothing algorithm, called GRIPS, that deforms given trajectories by locally optimizing the placement of vertices while satisfying the system's kinodynamic constraints. We show experimentally that GRIPS typically produces trajectories of significantly higher smoothness and smaller length than several existing post-smoothing algorithms.

If using GRIPS for scientific publications, please cite the following paper:

@inproceedings{heiden2018grips,
  author={Heiden, Eric and Palmieri, Luigi and Koenig, Sven and Arras, Kai O. and Sukhatme, Gaurav S.},
  booktitle={IEEE International Conference on Robotics and Automation (ICRA)},
  title={Gradient-Informed Path Smoothing for Wheeled Mobile Robots},
  year={2018}
}

Requirements

• CMake >=3
• Eigen 3
• OMPL ~1.3.1
• Qt5 (ensure the Qt Charts and SVG packages are installed)

Content

The following CMake targets are available:

CMake target	Description
homotopy_test	Compares paths from Theta* and A* before/after post-smoothing w.r.t homotopy class
benchmark	Compares different post-smoothing and path planning algorithms (cf. Table 1) and generates statistics JSON in log folder
shortening_test	Compares path-shortening results on hand-crafted path (Fig. 2)
showcase	Visualizes post-smoothing of Theta* path in S-shaped environment (Fig. 3)

Used third-party tools

• nlohmann/json to store path statistics data for plotting
• palmieri/posq steer function for differential drive robots