Online verification of multiple safety criteria for a robot trajectory

Abstract

Abstract-Ensuring the safety of humans in a collaborative environment with industrial robots is a major concern of human-robot co-working. Most current approaches give no formal guarantee of safety; where such guarantees are given, accounting for the unpredictability of the human may limit robot efficiency. We therefore developed a novel trajectory planner for a robot arm, which formally guarantees the safety of humans from collision with the robot for every possible human behaviour, without restricting the robot more than necessary. To achieve this, the trajectory planner verifies online that the robot adheres to two separate safety criteria derived from ISO standards using reachable occupancies of surrounding humans and the robot. While one criterion anticipates all possible movements of the human and provides a reasonable safety guarantee, the other criterion additionally guarantees strict safety as long as the human behaves as expected according to ISO standards. We implemented the trajectory planner for a real robot arm and show some experimental results.

Safe Trajectory Tracking in Uncertain Environments

“…The MPFTC framework builds on the assumption that a safe set exists, where all (including the unknown) constraints are satisfied at all time: this is an assumption often made in practice for stable systems at rest, as long as a safe configuration can be found. E.g., a stopped car, an empty reservoir, a switched off electric circuit, etc [14], [15]. Similar arguments apply to controlled unstable plants.…”

“…Lemma 4. Assume that g n|k is defined according to (15) and Assumption 9 holds, then Assumption 7 is satisfied.…”

“…Indeed, optimality would require M = ∞; with a finite M we therefore accept some degree fo suboptimality, as is customary in the formulation of terminal conditions for MPC. By expressing the a priori unknown constraints as in (15), it is clear that if γ i n|k (x n|k , u n|k , w n|k ) > 0, for some i, then Problem (4) becomes infeasible. Since the argument of safety relies on Assumption 10, we ensure that whenever γ i n|k (x n|k , u n|k , w n|k ) > 0, component i of γ is removed, and an additional component j is added to g as…”

Enhancing fluency and productivity in human-robot collaboration through online scaling of dynamic safety zones

“…In [19], bounding volumes as sphere swept lines (SSLs) are implemented to compute online proximity checks between the robot and its human co-worker. More recently, the SSLs are adopted by the authors in [20] to guarantee safety by considering path-consistent trajectories for every possible human behavior. In [21], the probability of human hand intrusion in the robot workspace is estimated using interference theory applying the SSM safety method.…”

Online Computation of Time-Optimization-Based, Smooth and Path-Consistent Stop Trajectories for Robots

“…In order to achieve path consistency, the authors of [21][22][23][24][25][26] implemented different techniques inside the control loop to dynamically choose the linear scaling of the desired trajectory. Multiple safety criteria are also considered in [27] and verified online to formally ensure path-consistent fail-safe stops. Other applications of path-consistent stopping trajectories from the perspective of optimal control have been proposed for autonomous vehicles in, e.g., [28,29].…”