Differentiable Gaussian Process Motion Planning

Modern trajectory optimization based approaches to motion planning are fast, easy to implement, and effective on a wide range of robotics tasks. However, trajectory optimization algorithms have parameters that are typically set in advance (and rarely discussed in detail). Setting these parameters properly can have a significant impact on the practical performance of the algorithm, sometimes making the difference between finding a feasible plan or failing at the task entirely. We propose a method for leveraging past experience to learn how to automatically adapt the parameters of Gaussian Process Motion Planning (GPMP) algorithms. Specifically, we propose a differentiable extension to the GPMP2 algorithm, so that it can be trained end-to-end from data. We perform several experiments that validate our algorithm and illustrate the benefits of our proposed learning-based approach to motion planning.

Authors

• Mohak Bhardwaj
• Byron Boots
• Mustafa Mukadam

Keywords

• Planning
• Conferences
• Automation
• Gaussian processes
• Artificial intelligence
• Trajectory optimization
• Robots
• Gaussian Process
• Path Planning
• Wide Range Of Tasks
• Planning Algorithm
• Linear System
• Likelihood Function
• Feed-forward Network
• Nonlinear Programming
• Nonlinear Least Squares
• Complex Distribution
• Maximum A Posteriori
• Training Module
• Tight Constraints
• Least-squares Optimization
• Hinge Loss
• Factor Graph
• Signed Distance Function
• Mixed Dataset
• Robots In Environments
• Planning Module
• Backpropagation Through Time
• Collision-free Trajectory
• Velocity Limits
• Occupancy Grid
• Robot Motion
• Neural Network
• Distribution Problem
• Computational Graph
• Dirac Delta