Trusting the Process

For my UC Berkeley Master of Engineering capstone, I worked with the Co-Design Lab on a project exploring trust in human-robot interaction. We built a configurable Unity-based simulation of a robotic pick-and-place task, where participants could intervene when they stopped trusting the robot.

I helped build the simulation and its robot-object interactions, including end-effector pickup and drop behavior, multi-joint motion, and configurable smooth or jagged robot movement. We used intervention time as a behavioral measure of trust and ran preliminary trials to investigate whether awkward robot motion caused people to intervene sooner.

The resulting platform was designed as a general-purpose testbed that future researchers could extend to study trust and human-robot collaboration in more realistic tasks.