11/07/2025
By Danielle Fretwell

The Francis College of Engineering, Department of Electrical and Computer Engineering, invites you to attend a Master's Thesis defense by Nathan Uhunsere on: "Assessing a Generalized Physical Testbed for Human Robot Teaming Studies."

Candidate Name: Nathan Uhunsere
Degree: Master’s
Defense Date: Friday, November 21, 2025
Time: 2 - 2:30 p.m.
Location: Ball 313

Committee:

  • Advisor: Paul Robinette, Associate Professor, Electrical and Computer Engineering, University of Massachusetts Lowell
  • Maru Cabrera, Assistant Professor, Miner School of Computer & Information Sciences, University of Massachusetts Lowell
  • Orlando Arias, Assistant Professor, Electrical and Computer Engineering, University of Massachusetts Lowell

Abstract:
Experiments that aim to study the dynamics of human-robot teams are often performed with scalable low-fidelity virtual environments. Physical environments, while seen as being more ecologically valid, are frequently prohibitive to construct and lack generality. This thesis proposes a novel framework that would help ease these concerns; providing justification and an outline for a physical testbed that is accessible, generalizable, and maintains ecological validity.

To investigate the justification for physical testbeds an experiment is constructed (N = 35), in which participant responses across parallel virtual and physical experimental environments are compared. In these environments participants will be asked to play through a simplified version of the game "Overcooked" once in a low-fidelity virtual environment (created in PyGame) and another in a physical environment with the Boston Dynamics Spot robot.

The study will examine the changes in metrics such as mental workload, moral trust and performance trust. The findings aim to quantify the differences between virtual and physical experiments, testing the general justification for using high-fidelity physical testbeds over low-fidelity virtual ones. This will in turn validate the construction of a generalizable framework to enable more scalable yet ecologically valid HRI research.