Improving area coverage by reversible object pushing

This article presents the implementation of an innovative safety module for a robot control architecture. It applies the principle of reversibility to assess intrinsic safety of actions and to adapt robot´s behavior. The underlying idea is that all reversible actions are intrinsically safe. A practical experiment is conducted to demonstrate the approach. The robot´s task is to cover a given area with a movable object inside. A governor, acting upon the principle of action reversibility, allows pushing only if such an action is reversible. It is compared with two other governors - one always allows all the actions and the other allows only actions that will not move the object. The covered area and changes between initial and final position of the object are measured in the end of each test run. The proposed governor allows full area coverage with minimal change in object position, while other governors exhibit either incomplete area coverage or significant change in object position. This is interpreted as the ability of the proposed reversibility-based approach to generate both effective and safe robot behaviors.