A dynamical system approach to task-adaptation in physical human–robot interaction

Abstract

The goal of this work is to enable robots to intelligently and compliantly adapt their motions to the intention of a human during physical Human-Robot Interaction (pHRI) in a multi-task setting. We employ a class of parameterized dynamical systems that allows for smooth and adaptive transitions between encoded tasks. To comply with human intention, we propose a mechanism that adapts generated motions (i.e., the desired velocity) to those intended by the human user (i.e., the real velocity) thereby switching to the most similar task. We provide a rigorous analytical evaluation of our method in terms of stability, convergence, and optimality yielding an interaction behavior which is safe and intuitive for the human. We investigate our method through experimental evaluations ranging in different setups: a 3-DoF haptic device, a 7-DoF manipulator and a mobile platform.Keywords Physical human-robot interaction · Adaptive behavior · Compliant control · Dynamical systems · Predictive models All authors are with the Learning Algorithms and System-

Differential game theory for versatile physical human–robot interaction

“…How should a contact robot be controlled to provide a stable and appropriate response to a user with unknown dynamics during various activities ranging from sport training, to physical rehabilitation and shared driving [8,9,10]? Specific human-robot interactions have been studied [11,12] but a general framework for interactive control is still missing. It has been suggested that differential game theory (GT) can be used as a framework to describe various interactive behaviors between a robot and its human user [13].…”

Human-Humanoid Collaborative Carrying

“…Methods combining machine learning and dynamical systems, e.g. [58] could be investigated as they could also encompass the planning part.…”

Force Adaptation in Contact Tasks with Dynamical Systems

“…In the second scenario, we perform a collaborative task with a human where the human asks the robot to clean the surface at different locations. To achieve this, we combine the proposed force adaptation with a mechanism to adapt the attractor of a nominal limit cycle (proposed in our previous works [22], [31]). We show that the force modulation can adapt fast enough to cope with the change in dynamics.…”

“…• The work in [31] to switch across different tasks. For the cleaning of the surface we define two tasks: -The homing task (i = 1) defined by f 1 (x) = x a,h − x and F 1 d (x) = 0 ∀x, where the robot should reach a fixed attractor above the surface, with x a,h the attractor.…”

“…T is the belief vector whose update rule ensures that b 1 + b 2 = 1 and that one b i → 1 in finite time (see [31]). Concerning the overall desired force profile F d (x), we set it to:…”