The Haptic Shoulder for Physical Human Robot Interaction project, sponsored by Elizabeth Peiros at the UCSD ARCLab, studied the range-of-motion (ROM) of the human shoulder's 3 rotational degrees of freedom (DOF) during upper-arm movement. The human shoulder experiences resistance when pushed towards its end ROM and this resistive force changes with the displacement of the upper arm. Specifically, the project focused on the motion, named upper-arm repositioning (UAR), that is commonly performed for search-and-rescue (SAR) operations.
In human casualty recovery situations, the victim is often found in a non-ideal configuration and needs to be repositioned for extraction. It is imperative that this is done safely without injury to the victim. Current methods of extraction by a robot have not heavily explored the first-step of repositioning a body for a safe extraction. Common steps for casualty repositioning are upper-arm repositioning and logrolling. These steps are performed to maintain spinal alignment and prevent further injuries and complications.
The focus of the project was to design a device that could quantify the resistive forces and torques produced in the shoulder. A human test subject can only provide qualitative data about the forces they “feel” and experiments are limited by the availability of willing test subjects. The project's motivation was to design a training tool for an existing robot arm that would enable the robot to learn how to safely perform upper-arm repositioning across multiple use-cases, without requiring the use of a human test subject.
Major requirements of the device included:
Mechanical – achieve accurate 3DOF motion using an effective singular joint center with a maximum separation of 2.5 cm (1 in)
Sensing – determine when the shoulder reaches a specified area in 3D space
Electrical – apply detectable, variable, and proportional resistive torque during end ROM
Important results of the project:
Device achieve proper ROM for upper-arm repositioning
Device replicated accurate human shoulder motion, i.e. singular joint center
Coupled motors provide variable ROM
Stiffness constant adjusted in code for multiple-use testing
The SHULDR Device satisfies all assigned performance requirements and is capable of being expanded upon for other experiments that are beyond the scope of this project. The final design incorporates two distinct joints and an electrical system that can be set to varying stiffnesses. The coupled motors allow the testing of different flexibilities and upper arm configurations.
Figure 1. Final design of the SHULDR Device
Figure 2. Final CAD of the SHULDR Device