Machine Learning Intersection Timing Tool: Analysis of New York City’s Midtown in Motion Initiative; On Opposite Sides of the Road: Examining the Discourse Surrounding Driving and Other Transportation Systems in America Utilizing Geels’ Multi-Level Perspective

Somehow finding yourself sitting behind your 8th red light on a one mile trip is by far one of the most frustrating experiences I have had the displeasure of enduring. That and my general experience with driving lead me to explore transportation in America. Transportation and logistics is an incredibly important part of any nation, yet so often seems to be an afterthought or just simply poorly managed. Spawning from my annoyance with current systems I began to look into methods of improving current systems. In my technical portion I proposed a machine learning based traffic signal optimizer to avoid those perpetual red lights, then researched other “smart” road improvements such as New York City’s “Midtown in Motion”, and finally possible alternatives such as a transition to public transit first infrastructure.
The technical portion of my thesis proposes a web application utilizing machine learning to optimize traffic light timings. Through this tool municipalities and engineers could simply enter in a traffic light that they would like to be optimized, then the program would collect publicly available data on car position, speed, etc at that traffic light over the course of the past year. A machine learning algorithm, specifically reinforcement learning, would then use the data to determine the optimal timings for that intersection. This tool would allow engineers and municipalities to input data about intersections, generate efficient light timings, and improve traffic throughput. Hopefully addressing the inefficiencies of outdated signal timings across the United States that cause significant delays, emissions, and alleviating my personal frustrations.
My subsequent STS analysis investigates the “Midtown In Motion” initiative, an intelligent traffic control system implemented in NYC to alleviate congestion in midtown Manhattan. MiM evaluates the system's impacts and explores interpretive flexibility, privacy concerns, and its role in the broader transportation debate. The Social Construction of Technology (SCOT) lens also questions the closure of car-centric infrastructure and considers the potential redefinition of urban transportation priorities.
In my STS research, I examine the challenges of America’s car-centric transportation system, highlighting its significant health, environmental, and economic costs. Public transportation has clear benefits such as reduced emissions, improved health, and increased equity. Despite public transit benefits and driving’s contribution to rising obesity rates, mental health issues, and high carbon emissions, government spending disproportionately favors highways over public transit. My STS research is founded on the fact that fragmented discourse and inconsistent support have hindered meaningful reform. After analyzing the available discourse through Geels’ multi-level perspective, I found the potential for change as younger generations are less car-dependent and more environmentally conscious causing a shift in cultural and political priorities. Secondly, by reframing public transit as a public good that enhances health, sustainability, and connectivity, policymakers can align fragmented arguments and foster transformative progress. Achieving unified support across societal, political, and technological levels is crucial to creating a healthier, more equitable, and sustainable transportation system for the future.
The use of a sociotechnical perspective emphasizes the need for engineers to consider the broader societal impacts of their work, aligning technical success and efficiency with goals like reducing emissions and promoting equity. For example, my proposed traffic optimization tool and the broader push for public transit-first infrastructure are most effective when framed as contributions to public health, environmental sustainability, and social connectivity. Ultimately, this perspective advocates for a holistic approach to transportation reform in an effort to avoid “thintelligence” and by recognizing that sustainable progress requires coordination across technical, social, and political dimensions. By integrating STS principles, engineers can contribute to systems that prioritize ethical responsibility, inclusivity, and long-term societal benefit.

School of Engineering and Applied Science

Bachelor of Science in Computer Science

Technical Advisor: Brianna Morrison

STS Advisor: Kathryn Neeley

Technical Team Members: Saidamir Umarkhodjaev