Real-time Batched Distance Computation for Time-Optimal Safe Path Tracking

TL;DR

This paper proposes a batched, fast and precise distance checking method based on precomputed link-local SDFs that can check distances for 500 waypoints along a trajectory within less than 1 millisecond using a GPU at runtime, making it suited for time-critical robotic control.

Abstract

In human-robot collaboration, there has been a trade-off relationship between the speed of collaborative robots and the safety of human workers. In our previous paper, we introduced a time-optimal path tracking algorithm designed to maximize speed while ensuring safety for human workers. This algorithm runs in real-time and provides the safe and fastest control input for every cycle with respect to ISO standards. However, true optimality has not been achieved due to inaccurate distance computation resulting from conservative model simplification. To attain true optimality, we require a method that can compute distances 1. at many robot configurations to examine along a trajectory 2. in real-time for online robot control 3. as precisely as possible for optimal control. In this paper, we propose a batched, fast and precise distance checking method based on precomputed link-local SDFs. Our method can check distances for 500 waypoints along a trajectory within less than 1 millisecond using a GPU at runtime, making it suited for time-critical robotic control. Additionally, a neural approximation has been proposed to accelerate preprocessing by a factor of 2. Finally, we experimentally demonstrate that our method can navigate a 6-DoF robot earlier than a geometric-primitives-based distance checker in a dynamic and collaborative environment.

Figures (12)

• Figure 1: Problem Setting Overview: Computing distances in real-time, across multiple robot configurations, with precision. See \ref{['sec:intro']} for more information.
• Figure 2: A common way to compute a distance between pointcloud and SDFs requires pointcloud transformations for each link coordinate. Instead, we transform and merge the link-local SDFs into the global coordinates in a pre-processing stage, and then evaluate it to obtain distances at runtime.
• Figure 3: A pipeline of parallel batched distance checking with pre-computed link-wise signed distance fields (SDFs). This is in 2D for clear illustration, but actual computation is in 3D and in a batched manner. See \ref{['sec:our_method']} for detail.
• Figure 4:
• Figure 5: Computation speed of euclidean grid approximation by a tiny neural net. See \ref{['subsec:neuralnet_approximation']} for detail.