The estimation of grip force for surgical tools such as the da Vinci has been shown to be valuable in possible applications such as haptics, tissue identification, and surgical training. Successful estimation attempts have been previously demonstrated, but utilize customized sensors; this letter aims to provide an estimate considering only typical sensor streams already present in commercially available surgical robots. The objective of this letter is to evaluate three proximal-end torque surrogate methods in their abilities to estimate distal-end states. The estimates are compared with previously reported results found in literature and the percent difference between the customized sensor approach and previous standards is reported. The most effective surrogates for proximal-end torque were commanded motor current and measured motor current. The jaw angle estimate resulted in 0.37 degree root mean square error, and the distal-end torque estimate resulted in 4.42 mNm RMSE, which compares favorably to existing literature approaches.
- AI-based methods
- Surgical robotics: Laparoscopy
- medical robots and systems
- perception for grasping and manipulation