Abstract
Thesis Project Portfolio
Intelligent Roadway Insight System (IRIS)
(Technical Report)
Under Watch: How Driver Monitoring Systems Reshape Labor Relations
(STS Research Paper)
An Undergraduate Thesis
Presented to the Faculty of the School of Engineering and Applied Science
University of Virginia • Charlottesville, Virginia
In Fulfillment of the Requirements for the Degree
Bachelor of Science, School of Engineering
Dimitris Veliadis
Spring, 2026
Department of Electrical Engineering
Table of Contents
Executive Summary
Intelligent Roadway Insight System (IRIS)
Under Watch: How Driver Monitoring Systems Reshape Labor Relations
Prospectus
Executive Summary
Road safety in the United States remains a major public problem, and fatigue among
commercial drivers is one important part of it. In 2023, drowsy driving contributed to over
300,000 police-reported crashes, more than 100,000 injuries, and about 6,400 fatalities. Large
trucks were involved in about 10% of all traffic fatalities, and more than 80% of those killed in
large-truck crashes were people outside the truck. At the same time, long-haul truck drivers may
still work more than 60 hours per week under federal hours-of-service rules, meaning fatigue is
not simply an individual failure but a recurring condition shaped by schedules, delivery
demands, and the organization of freight labor. One common response has been the spread of
electronic logging devices and driver monitoring systems (DMS) that record driver behavior and
attempt to reduce crash risk through alerts and compliance tracking. Yet these systems do not
fully address the structural causes of fatigue, and they can intensify surveillance by shifting
attention toward driver behavior rather than toward working conditions. My thesis portfolio
addresses this broader problem of how fatigue can be reduced in trucking without simply
expanding intrusive monitoring. The technical project approaches the issue through privacy-
conscious device design, while the STS research examines how existing monitoring systems
reshape labor relations, privacy, and responsibility in the trucking industry.
My technical project, Intelligent Roadway Insight System (IRIS), explores how a fatigue-
detection device can improve safety while preserving driver privacy. IRIS is a wearable system
in the form of a baseball cap designed primarily for long-haul truck drivers. Instead of relying on
employer-controlled in-cab cameras, the device keeps sensing and processing on the wearer
through edge AI. The cap integrates a camera for eye-closure monitoring, an accelerometer for
head-motion detection, and a heart-rate sensor for psychological data, all processed through a
MAX78000 microcontroller with an on-chip neural network accelerator. These signals are fused
on-device to classify drowsiness and trigger a haptic alert when dangerous behavior is detected.
The final prototype successfully demonstrated the core sensing and processing pipeline using
evaluation boards mounted on a cap, and the eye open/closed classifier exceeded the project
target of 75% accuracy. The prototype also showed that on-device fusion of eye, motion, and
heart-rate data could be used to drive real-time alerts without requiring continuous transmission
of personal data. At the same time, the project did not fully achieve all design goals. Most of the
system still relied on evaluation boards rather than the intended custom PCB network, the lack of
infrared illumination limited reliability at night, and the camera placement partially obstructed
one eye’s forward view. Even with these shortcomings, the project demonstrated that a more
privacy-conscious alternative to conventional in-cab monitoring is technically feasible and worth
further development.
My STS research paper, Under Watch: How Driver Monitoring Systems Reshape Labor
Relations, examines how DMS technologies change truckers’ working conditions and the way
responsibility for fatigue is assigned. The project asks how the implementation of DMS has
reconfigured labor relations around fatigue management. Using qualitative, document-centered
research, I analyzed three main categories of evidence: policy and regulatory materials, vendor
and industry materials, and scholarly research on fatigue, monitoring, and workplace
surveillance. I also used the frameworks of mutual shaping and Helen Nissenbaum’s contextual
integrity to evaluate how monitoring technologies both respond to and reshape trucking labor
conditions, and to assess whether the information flows created by DMS are appropriate to the
context of trucking work. The paper argues that fatigue in trucking is structurally produced by
irregular schedules, time pressure, and limited rest feasibility, but DMS often reframes that
1
structural problem as an individual behavioral issue that can be tracked, scored, and reviewed.
Vendor materials present DMS as an objective safety solution, while broader research shows that
these systems can also support discipline, liability management, and expanded workplace
surveillance. My conclusion is that DMS does not simply detect fatigue; it redistributes control
and responsibility in ways that often shift blame toward individual drivers while leaving the
deeper causes of fatigue intact. In that sense, monitoring technologies may reduce some
immediate risks, but they also create new privacy and labor-governance concerns that must be
taken seriously.
2
