Project Overview:

This project sponsored by Elizabeth Peiros at the UC San Diego ARCLab focuses on the quantitative measurement of reaction torques in the human shoulder during its end range-of-motion (ROM). The sponsor trains a robotic manipulator to safely reposition a human body for extraction during search and rescue operations. It is important to reorient a human body for safe extraction and to avoid damaging soft tissue. This project’s motivation was to find an alternative way for the sponsor to safely train the robot without using their own body and to have multiple-case scenarios. The solution was to design a shoulder mannequin that can measure when the upper arm reaches its end range of motion and apply a reaction torque in response. The complete system consists of a complex joint that moves with three degrees of freedom (DOF), and an electrical system that applies torque. The major requirements for the mannequin included achieving the necessary ROM to perform upper-arm repositioning and applying a resistive reaction torque during the end ROM. The delivered product surpasses the ROM needed for upper-arm repositioning and can output a measurable force at the grasping point.

My contributions:

I worked on a novel universal joint (U-joint) designed to accurately mimic human shoulder motion for robotic rescue applications, focusing on key aspects such as joint center, mass distribution, cable management, compliance, and kinematic analysis. This design incorporated motors and encoders, simplifying software controls while ensuring anatomically precise shoulder movements. The challenge was to replicate the full range of motion (ROM) of a human shoulder with a singular center of rotation (COR), a difficult balance to achieve in one design. After evaluating various concepts using a Pugh chart and prototyping two selected designs, I combined the best features from each. Key modifications included inverting the small basket orientation to improve weight counterbalance and adding a wedge to the joint base to enhance the ROM, particularly for upper arm repositioning. These refinements ensured the final design was both functional and accurate, making it a robust component for advanced human-robot interaction.