Litcius/Paper detail

Collision Detection, Identification, and Localization on the DLR SARA Robot with Sensing Redundancy

Maged Iskandar, Oliver Eiberger, Alin Albu‐Schäffer, Alessandro De Luca, Alexander Dietrich

202137 citationsDOIOpen Access PDF

Abstract

Physical human-robot interaction is known to be a crucial aspect in modern lightweight robotics. Herein, the estimation of external interactions is essential for the effective and safe collaboration. In this work, an extended momentum-based disturbance observer is presented which includes the sensing redundancy related to additional force-torque measurements. The observer eliminates the need for acceleration measurements/estimates and it is able to accurately reconstruct multiple simultaneous contact locations. Moreover, it provides uncoupled, configuration-independent, and singularity-free estimates of the external forces. The performance of the approach is experimentally validated on the SARA robot, the new generation of DLR lightweight robots, involving high resolution force-torque sensors in a redundant arrangement.

Topics & Concepts

Redundancy (engineering)RobotTorqueComputer scienceAccelerationArtificial intelligenceRoboticsContact forceObserver (physics)Control theory (sociology)CollisionControl engineeringSimulationComputer visionEngineeringPhysicsComputer securityThermodynamicsQuantum mechanicsClassical mechanicsControl (management)Operating systemRobot Manipulation and LearningTeleoperation and Haptic SystemsSoft Robotics and Applications