Researchers have developed a novel model-based system for estimating grasp forces in robotic grippers using visual feedback from RGB-D cameras. This approach integrates iterative contact localization with an inverse finite element analysis simulation, allowing it to generalize to unseen objects and conditions. The system demonstrated high accuracy, with an average root mean square error of 0.23 N during the load phase and 4.34% over the entire grasping process. AI
Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →
IMPACT Improves robotic manipulation safety and control by enabling accurate indirect force sensing.
RANK_REASON Academic paper detailing a new system for robotic grippers.