Answers are grounded in this paper and cite their sources.
Shows improved contact force estimation in human-robot collaboration, highlighting efficiency and accuracy.
This paper presents a sensorless method for estimating both contact position and contact force without relying on additional hardware. The proposed approach formulates the contact estimation problem as an optimization task and introduces a scalar equivalent contact parameter to simplify the process, reducing the number of decision variables from seven, as required by conventional methods, to a single scalar. This reduction significantly decreases the risk of convergence to local minima while improving robustness and computational efficiency. External joint torques are first estimated using the robot dynamic model together with encoder feedback. The equivalent contact parameter is then solved to determine the equivalent contact point, from which the corresponding equivalent contact Jacobian is derived. The contact force is subsequently estimated using the external joint torques and the equivalent Jacobian. Finally, the actual contact point is obtained by intersecting the estimated force line of action with the robot geometric surface model. The proposed method is validated through simulations and experiments on a commercial heavyweight industrial robot. The results demonstrate superior performance over conventional techniques in both computation time and estimation accuracy, confirming the method's effectiveness and practical applicability in human–robot collaboration.
No takes yet. Share an insight, caveat, or question.
Ta et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: