Robot Collisions: A Survey on Detection, Isolation, and Identification

Abstract

Abstract-Robot assistants and professional co-workers are becoming a commodity in domestic and industrial settings. In order to enable robots to share their workspace with humans and physically interact with them, fast and reliable handling of possible collisions on the entire robot structure is needed, along with control strategies for safe robot reaction. The primary motivation is the prevention or limitation of possible human injury due to physical contacts. In this survey paper, based on our early work on the subject, we review, extend, compare, and evaluate experimentally model-based algorithms for realtime collision detection, isolation, and identification that use only proprioceptive sensors. This covers the context-independent phases of the collision event pipeline for robots interacting with the environment, as in physical human-robot interaction or manipulation tasks. The problem is addressed for rigid robots first, and then extended to the presence of joint/transmission flexibility. The basic, physically-motivated solution has already been applied to numerous robotic systems worldwide, ranging from manipulators and humanoids to flying robots, and even to commercial products.

A model‐based strategy for safety assessment of a robot arm interacting with humans

“…Fig. There are several known contact detection schemes a robot can use, see [3], but the reaction mechanism is usually robot specific. It can be seen that by changing the contact point, the measured contact force changes not just quantitatively but also qualitatively.…”

“…1 by applying presented strategy for safety assessment, the model must be extended to also include the contact detection and robot reaction mechanism. There are several known contact detection schemes a robot can use, see [3], but the reaction mechanism is usually robot specific. If these mechanisms are not known or can not be identified, an alternative solution would be to exploit all knowledge of the physics combined with machine learning approaches to estimate the unknown processes.…”

Soft-bubble: A highly compliant dense geometry tactile sensor for robot manipulation

“…In addressing the need to sense and manage physical interactions, there are several themes that have been tackled so far: sensing accidental contacts and reacting suitably [10], tracking and controlling contact forces during intentional contact [11], tactile exploration for objects that are hardto-see or occluded by clutter [3], classification of object type and shape inference [12] and sensing the quality of induced grasps [13]. For tactile sensing of object class (shape, material, etc.)…”

Interval-Arithmetic-Based Trajectory Scaling and Collision Detection for Robots with Uncertain Dynamics

“…If the current exceed this threshold, they can distinguish between an intentional contact or an accidental collision, for which they have different reaction strategies. A recent survey about collision detection is presented in [13]. There, the authors compare various collision detection schemes and list their respective advantages and disadvantages.…”