Reachable Sets for Safe, Real-Time Manipulator Trajectory Design

Patrick Holmes, Shreyas Kousik, Bohao Zhang, Daphna Raz, Corina Barbalata, Matthew Johnson, Ram Vasudevan


For robotic arms to operate in arbitrary environments, especially near people, it is critical to certify the safety of their motion planning algorithms. However, there is often a trade-off between safety and real-time performance; one can either carefully design safe plans, or rapidly generate potentially-unsafe plans. This work presents a receding-horizon, real-time trajectory planner with safety guarantees, called ARMTD (Autonomous Reachability-based Manipulator Trajectory Design). The method first computes (offline) a reachable set of parameterized trajectories for each joint of an arm. Each trajectory includes a fail-safe maneuver (braking to a stop). At runtime, in each receding-horizon planning iteration, ARMTD constructs a parameterized reachable set of the full arm in workspace and intersects it with obstacles to generate sub-differentiable, provably-conservative collision-avoidance constraints on the trajectory parameters. ARMTD then performs trajectory optimization over the parameters, subject to these constraints. On a 6 degree-of-freedom arm, ARMTD outperforms CHOMP in simulation, never crashes, and completes a variety of real-time planning tasks on hardware.

Live Paper Discussion Information

Start Time End Time
07/16 15:00 UTC 07/16 17:00 UTC

Virtual Conference Presentation

Supplementary Video

Paper Reviews

Review 3

Overall, the work is of high quality and the results are impressive, but the paper is not matching the quality of the work. While the first part is clear and easy to follow, there are some notions that are not clearly motivated and explained. For instance the slicing is introduced early in the paper but its objective only explained later. As the paper introduces many notions, it is hard to get the global picture of how everything works together. I would recommend producing a figure detailing the offline and the online step to provide a gentler introduction to the method. A particularly annoying point concerns the supplementary document that seems to be simply a previous version of the paper. This strikes me as a demonstration of laziness; how hard was it to simply extract the proofs and relevant information from the document and present them on a reduced file ? The sections are not even properly refered in the paper. This is clearly not acceptable. This being said, again I think the work is of high quality.