Estimating the Maximum Pushing Force of Robot Manipulators to Physically Assist Caregivers

Abstract

Robotic use cases where a robot must perform a force application can lead to overloaded joints due to excessive external joint forces. This results in performance degradation or even hardware damage in the worst case. Such cases are conceivable, for example, in physical human-robot interaction, where the collaborating robot should remain relatively small and light, but large loads have to be handled several times a day in collaboration with humans, e.g., in nursing care. For this purpose, we present an approach to estimate suitable configurations of a predetermined task with a start and goal robot pose to generate a robot manipulator path which possesses suitable conditions to leverage the manipulator's maximum pushing force potential. We also evaluate different metrics to show that robot base placement plays a major role when utilizing the robot manipulator's pushing capability for different tasks. Based on the robot's configuration and/or placement, the pushing force potential can differ by several magnitudes.