Dynamically Constrained Motion Planning Networks for Non-Holonomic Robots
This work addresses the problem of reliable and fast planning for kinematically constrained robots in automated environments, representing an incremental improvement over existing neural planning methods.
The paper tackles real-time motion planning for non-holonomic robots by proposing Dynamic MPNet, an extension to Motion Planning Networks, which improves data efficiency and generalizability in simulation and an indoor navigation task with a Dubins car.
Reliable real-time planning for robots is essential in today's rapidly expanding automated ecosystem. In such environments, traditional methods that plan by relaxing constraints become unreliable or slow-down for kinematically constrained robots. This paper describes the algorithm Dynamic Motion Planning Networks (Dynamic MPNet), an extension to Motion Planning Networks, for non-holonomic robots that address the challenge of real-time motion planning using a neural planning approach. We propose modifications to the training and planning networks that make it possible for real-time planning while improving the data efficiency of training and trained models' generalizability. We evaluate our model in simulation for planning tasks for a non-holonomic robot. We also demonstrate experimental results for an indoor navigation task using a Dubins car.