Abstract
Modern governance has shifted from traditional power structures to systems rooted in data collection and surveillance. The technical project analyzes the University of Virginia (UVA) fleet to recommend electrification using total cost of ownership (TCO) and sustainability goals. The project specifically uses highly granular data and each vehicle’s specifications and lifecycle costs to analyze and recommend electrification based on vehicle efficiency, TCO, and greenhouse gas emissions (GHGs). The sociotechnical report examines how surveillance technologies have steered away from its original purpose based on foundational surveillance narratives into a more nuanced interpretation and deliberately networked model of governing in order to gain power and exert control. With rapidly advancing technologies, not only have surveillance technologies increased in effectiveness and efficiency, but have also become societally normalized under the pretense of neutral or necessary, hiding the underlying political narrative behind choices and introducing implications on democratic accountability and civil liberties. The connection between the technical and sociotechnical project may not be as straightforward at first interpretation, but both underlying motivations can be traced back to understanding how any individual or group in power, whether that is political or at a university, has transitioned its power to instrument the increased data collection to enable control and predict behavior. This general idea of actors in power and enabling continuous observation and data collection for personal gain is not always negatively associated, but is undoubtedly quantifying data to instrument numbers into things that can be controlled and predictable.
The technical project works directly with UVA facilities management (FM) and their fleet manager to represent to their leadership that fleet electrification is a crucial strategy that should continue to be implemented to reduce GHGs and achieve long-term cost savings. UVA FM’s fleet provides a unique scenario to implement this electrification due to its structure as a university-style fleet, a geographically bounded, multi-use fleet. This project developed a data-driven decision-support framework that serves as an analytical tool to plan future fleets based on economic and environmental outcomes. This framework identified vehicles suitable for electrification and quantified trade-offs between cost and emissions in comparison with the potential replacement vehicle, typically an internal combustion engine (ICE) vehicle or plug-in hybrid electric vehicles (PHEVs). This TCO model is supported with a machine learning (ML) approach to predict fuel usage of replacement candidates with the new vehicle’s specifications with the old vehicle’s mileage data. This predicted fuel usage helps the model to support any potential replacements that have yet to be made with data. To test the TCO model, case studies were conducted using historical replacements made by UVA FM. The methodology of this technical project supports UVA FM's leadership objective of achieving the university’s sustainability goals in addition to the cost-benefit analysis.
The results of this project is primarily based on a retrospective audit of vehicle replacement decisions previously made by UVA FM. To produce accurate results that FM can utilize, the project considers that some vehicles that produce significantly high GHGs are not necessarily strong electrification candidates due to operational profiles, such as long idling periods. The case studies conducted in this project analyzed two replacement pairs. The first replacement case analyzes a replacement from an ICE vehicle to a full hybrid electric vehicle (FHEV). The second replacement case analyzes a replacement from an ICE vehicle to a battery electric vehicle (BEV). Both cases exhibit similar cost-saving results and CO2 emissions reduction, reinforcing UVA sustainability goals. The dual-component design of the framework that uses both a retrospective audit and a forward-looking ML prediction, reflects the practical needs of an institutional fleet manager. The audit component provides accountability by evaluating whether past replacement decisions delivered on projected savings, grounded in fuel and energy consumption data rather than manufacturer estimates. While the ML approach provides a tool for fleet managers to support future replacement decisions based on an accurately built model. Although this project is specified to be used under UVA FM’s fleet, it is applicable to other university fleets and the framework itself was designed for scalability.
In the sociotechnical paper, the research explores how private corporations and state institutions have begun to converge in surveillance technologies, all through the guise of necessity or neutrality. Specifically, the paper answers the question: How has the convergence of state and corporate surveillance transformed governance from targeted intelligence gathering to data-driven strategic monitoring, and what are the implications for democratic accountability and civil liberties? The rapid advancement of technology is increasing at a much higher rate than society’s capacity to fully understand and address its ethical implications. In recent years, the collaboration between public institutions and the private sector has extended past traditional surveillance infrastructures and has quickly placed these private corporations into a position of unprecedented power, and in many cases, introduced heightened concerns on accountability. The methodology of this paper approaches case studies and frames them around the convergence of governmentality and actor-network theory, where new actors have evolved in the government due to the rise of technologies and shift in surveillance narratives. The analysis conducted on the cases explains both why surveillance has become normalized and how power circulates through the network of actors, evaluating the political language, narrative framing, political context, and technological infrastructure involved.
The historical cases analyzed demonstrate the evolution of surveillance practices, beginning in the Cold War, where surveillance was justified through geopolitical fear and ideological conflict, and expanding during the post-9/11 War on Terror into large-scale data collection on civilians. Government programs such as NSA initiatives like Project Shamrock illustrate early forms of mass data collection and the resulting tensions with constitutional protections and democratic oversight. Whistleblower cases, including Edward Snowden’s disclosures, reveal hidden surveillance infrastructures and the involvement of private contractors. These dynamics persist in contemporary contexts, as seen in collaborations between Palantir Technologies and U.S. Immigration and Customs Enforcement (ICE), where corporate technology is integrated into governance to aggregate data and target individuals. Across these cases, political language consistently frames surveillance as necessary for national security and portrays technology as neutral, despite its clear political implications. Three patterns emerge: fear and geopolitical tension justify expanding surveillance; technologies become normalized as neutral systems; and corporate-state partnerships introduce powerful actors with limited accountability. As a result, power is no longer centralized within the state but distributed across interconnected networks of governments, corporations, and technologies. This data-driven, sociotechnical system enables granular data collection, increasing the capacity to profile, predict, and control behavior while normalizing reduced transparency and weakened democratic accountability.
