Abstract
This Undergraduate Thesis Portfolio brings together two bodies of work that, taken together, offer a fuller picture of automation than either could alone. The first is a technical Capstone project completed with the Virginia Department of Transportation (VDOT), in which I designed and deployed a low-code application to modernize an outdated government approval process. The second is an STS research paper that examines workplace automation as an ethical and political phenomenon, analyzing how automated systems redistribute power between workers, employers, and institutions. Though one project is primarily technical and the other primarily analytical, they share a common subject: automation and its consequences for human work. The Capstone demonstrates what automation can accomplish in practice, while the STS paper investigates what it means for the people it affects. Together, they represent an engineering education that takes seriously both the design of technology and the responsibility that comes with it.
The Capstone project addressed an inefficiency at VDOT, the manual, time-intensive process of preparing agreement approval forms for the Virginia Commissioner of Highways. Staff were spending an estimated 30 minutes per form navigating complex chain-of-command approval structures and entering data by hand, a process prone to errors that could restart the entire approval cycle and delay projects by days or weeks. To address this, I used Microsoft Power Apps and Power Automate to build an intuitive, low-code application that automatically generates pre-filled approval documents based on user inputs. The application integrates with SharePoint for document storage and draws on Microsoft 365 authentication to auto-populate user attributes such as name, title, division, and email. The result was a reduction in form completion time from roughly 30 minutes to under 5 minutes, along with significantly fewer submission errors and improved staff confidence in the accuracy of routed forms. The project demonstrated that low-code platforms can serve as practical, accessible tools for modernizing government administrative processes, freeing employees to focus on higher-value work that requires greater expertise.
The STS research paper investigates how workplace automation, while designed to enhance efficiency and reduce repetitive work, contributes to job displacement and shifting power dynamics between workers, employers, and technologies. The paper applies Langdon Winner’s Political Technologies framework, drawn from his 1980 essay “Do Artifacts Have Politics?”, to four case studies spanning six decades: the mechanical tomato harvester, self-checkout systems, autonomous vehicles, and low-code software platforms. Across each case, a consistent pattern emerges, which is that automation is rarely politically neutral. Design choices consistently favor those who fund and deploy automated systems while reducing the economic standing and agency of workers, particularly those already vulnerable. At the same time, the analysis acknowledges that automation’s consequences are not predetermined. When designed with transparency, meaningful human oversight, and participatory processes, automation can complement rather than displace human labor. The paper concludes that ethical automation requires engineers and policymakers to look beyond efficiency metrics and deliberately account for the distribution of power and the protection of worker agency.
Working on both projects simultaneously sharpened my thinking in ways that neither project could have done independently. Building the VDOT application gave me firsthand experience with the design decisions that automation requires, specifically choices about what to automate, how to handle errors, how to present logic to non-technical users, and how to earn trust from the people who will use the tool. Those practical decisions took on greater meaning as I read Winner, Holzer, and Autor for the STS paper. I found myself asking, during development, questions I might not otherwise have thought to ask: Who benefits from this routing logic? What happens to the staff who used to manage this process manually? Does automating this task free people for better work, or simply reduce their purpose? The STS paper gave me the language for those questions, and the Capstone gave me a real-life example through which to test the answers. In turn, the Capstone made the STS research feel less abstract. The case studies I analyzed were not just historical examples; they were versions of a dynamic I was participating in. That awareness made me a more careful and reflective engineer. The experience of holding both a technical and sociotechnical lens at once is, I think, the most valuable thing this portfolio represents. It goes beyond not just the ability to build automated systems, but the habit of asking whether and how they should be built.
