Experimental data complementing the paper "A Shared Control Approach to Safely Limiting Patient Motion Based on Tendon Strain During Robotic-Assisted Shoulder Rehabilitation" (IEEE ICORR 2025)

This dataset contains the experimental data collected during the experiments that were performed at the Cognitive Robotics department of TU Delft, to test the algorithms presented in the paper "A Shared Control Approach to Safely Limiting Patient Motion Based on Tendon Strain During Robotic-Assisted Shoulder Rehabilitation". The dataset, in form of ROS bags (recorded during the experiments with a human subject interacting with a KUKA robot controlled via impedance control in ROS 1) is made by:

1. "reactUnsafe_0.bag", containing the results of the low-authority shared controller presented in the paper. A portion of the data was employed to generate Figure 4 in the paper.
2. "deflection_interpTraj.bag", containing the results of the high-authority shared controller. A portion of the data is visualized in Figure 5.
3. "impact.bag" was recorded when the controller from a previous publication ("Enabling patient-and teleoperator-led robotic physiotherapy via strain map segmentation and shared-authority"), and supports the results in section III.B.

An example of how to extract the data from each .bag file can be found in the GitHub repository of the project (https://github.com/itbellix/biomechanical_safe_deflection) under scripts/plot_results.py