Tiago Breunig, Jerin Adkins, and Nathanson Gageby
“Tiago, Jerin, and Nathanson are students in WVU’s Critical Infrastructure Lab (Evansdale 2050), supervised by Dr. Mohamed Hefeida. Their winning project, “Beyond the Fixed Timer,” applies the lab’s AI-driven traffic modeling systems to design adaptive, energy-efficient signals for Appalachian roads. The project showcases how hands-on innovation in the Critical Infrastructure Lab leads to real-world solutions for smarter, safer communities.”
Beyond the Fixed Timer
Abstract
Traffic management at the crossroads of major highways and winding local roads presents a unique challenge across Appalachia. These critical intersections often rely on inefficient, fixed-timer signals, leading to unnecessary congestion, fuel waste, and driver frustration. This paper proposes a two-step modernization: first, updating these intersections with modern camera-based actuated controls, and second, using an innovative AI analysis pipeline to fine-tune the control algorithms for each specific location. This process allows for a tailored system that accounts for the unique traffic patterns of individual Appalachian intersections. This submission, based on a current WVU Capstone project called Evansdale 2050, demonstrates how this AI pipeline can create specialized, hyper-efficient traffic control systems, bringing intelligent, adaptive infrastructure to the region's roads.
Part I: The Optimal Solution – An AI-Enhanced Signal The Wait at the Crossroads
For thousands of West Virginia University students, the morning commute to the Evansdale campus is a daily test of patience, and the intersection of Patteson Drive and Morrill Way is the final hurdle. You inch forward in a line of cars, watching as the green arrow for the left turn onto campus stays lit just long enough for two or three vehicles to pass. This is often followed by a long green light for Morrill Way, a street with barely any exiting traffic that early in the morning. This frustrating imbalance is more than a minor inconvenience; it's a systemic inefficiency played out at countless crossroads across Appalachia.
The core issue is that these traffic systems are not smart—they are simply timed. They cannot see the long queue of students trying to get to class while Morrill Way sits empty, nor can they recognize the reverse during evening hours when traffic is trying to exit campus. To solve this traffic issue, Evansdale 2050 (a current WVU Capstone project) employs a unique development process centered on a primary traffic control algorithm. To perfect this algorithm, an AI is being used to analyze the algorithm and provide insights into improved algorithmic patterns. This observer AI watches the algorithm manage traffic in a custom-built lab model, identifying patterns and inefficiencies that the human eye might miss. Its analysis provides the data-driven input needed to constantly refine the core logic, allowing for the creation of a hyper-efficient model for the region's roads and proving that signals can be taught to see the traffic they are meant to control.
Forging a Smarter Signal: The Proposed Solution
Our approach begins by identifying and differentiating the complexity of the traffic in different areas of our region. An intersection like Patteson Drive and Morril Way, mentioned previously, is the intersection being observed and worked with in simulations to test and iterate on, which creates the opportunity for our project to implement a solution that can be scaled up or down to solve a similar problem in regions with different needs by implementing a modern, actuated traffic control algorithm as a foundation. Unlike the region's common fixed-timer signals, this system uses cameras to detect vehicles in real-time, allowing it to adapt to changing traffic flow. Furthermore, it can be programmed with schedules to prioritize specific lanes during predictable peak hours, such as a morning commute or when a local school lets out.
While this technology is a significant leap forward, its logic is still static; it follows pre-set rules and cannot adapt to the unique, unforeseen traffic patterns of a specific location. It is this potential for further optimization that our project addresses. We introduce an innovative AI pipeline that analyzes a simulation of the intersection running this baseline algorithm. The AI observes the flow, identifies hidden inefficiencies, and learns the unique 'rhythm' of that intersection—a rhythm a standard algorithm would miss. Based on this deep analysis, the AI refines and customizes the control algorithm. The result is a specialized system, hyper-efficient and tailored to the exact needs of one intersection. This process serves as a powerful demonstration of how AI can be integrated to solve complex, real-world infrastructure problems in Appalachia and beyond.
The Efficiency Revolution: Reclaiming Wasted Time and Resources
Once a foundation of safety is established, the conversation can turn to optimization. As discussed, a basic fixed-timer signal often trades safety for profound inefficiency, creating unnecessary delays that have tangible economic and environmental costs. This is where an intelligent, AI-actuated system provides revolutionary benefits.
Unnecessary idling is a quantifiable drain on resources. According to data from the U.S. Department of Energy, a typical passenger vehicle consumes between 0.2 to 0.5 gallons of fuel per hour while idling, depending on engine size and air conditioner use (U.S. Department of Energy). When you multiply that by hundreds or thousands of cars waiting at inefficient red lights across the state every day, the waste
becomes staggering. This translates directly to higher fuel costs for residents and businesses and an increase in carbon dioxide (CO₂) emissions. The data generated by simulations in our project aims to quantify the potential savings. By tailoring the signal's logic to the unique rhythm of its location, an AI-refined system can slash average wait times and, therefore, vehicle idle time.
Conclusion to Part I: A New Signal for Appalachia
This project demonstrates that the future of traffic management in Appalachia lies not in more concrete, but in smarter code. By using an AI observer to refine traffic control algorithms, it is possible to create a system that addresses the twin challenges of safety and efficiency, leading to quantifiable reductions in wasted fuel, frustrating delays, and dangerous intersection conflicts. This is not just a solution for a university town; it is a flexible model that can be scaled and adapted for the entire region. From the commuter corridors of the Eastern Panhandle to the small-town crossroads tucked in mountain valleys, this approach provides a cost-effective way to bring 21st-century infrastructure where it is needed most. Ultimately, Evansdale 2050 (a current WVU Capstone project) serves as a model for local innovation, proving that the region's brightest minds can solve its most persistent problems by creating a system that is not only safer and smarter, but finally allows its roads to be seen.
Part II: An Accessible First Step – The Intelligent Warning Beacon
While the AI-enhanced traffic signal represents the future of efficient roadway management, we recognize that its implementation requires significant infrastructure and investment. For many rural Appalachian intersections, a more immediate, lower-cost intervention is necessary to address the most pressing issue: safety. This section addresses that reality by proposing a pragmatic first step that uses foundational technology to reduce accidents at high-risk junctions where a full signal is not yet feasible.
The Safety Imperative: A Spectrum of Intervention
Many of Appalachia's most dangerous intersections are not in cities but on the rural highways connecting them. Picture a small local road, often obscured by hills or winding curves, ending at a two-way stop sign before it meets a 55-mph state highway. For a driver on the local road, pulling out requires judging the speed of distant, fast-approaching vehicles—a frequent cause of severe right-angle, or "T-bone," collisions.
Figure 2: Comparison of Fatalities in Signalized versus Unsignalized Intersections from he US Department of Transportation
The most fundamental safety enhancement for such a junction is not necessarily a full stoplight, but the installation of a simple warning beacon. A flashing yellow light suspended over the highway serves to jolt mainline drivers from complacency and signal an upcoming intersection, while a flashing red light facing the side road reinforces the stop sign. This beacon's primary role is to increase awareness and reduce accidents caused by inattention. However, this warning system does not eliminate the primary danger: the driver on the side road must still perform the life-threatening guesswork of judging gaps in high-speed traffic. While awareness is increased, the fundamental risk of a miscalculation remains.
For intersections with a documented history of severe crashes, even with a beacon, a more active intervention is required. The installation of a full traffic signal, which actively assigns right-of-way, is the definitive step in eliminating these dangerous conflict points by separating traffic flows in time. While our project's main focus is on making these signals intelligent, we must first acknowledge that a spectrum of solutions is needed, and simply providing active control at the most dangerous intersections provides an essential foundation of safety that is still lacking across the region.
Conclusion: A Strategic Vision for Appalachian Roads
The challenges facing Appalachian roadways are not uniform, and neither are the solutions. This paper presents a dual-pronged strategy that is both visionary and pragmatic, acknowledging that the path to safer and more efficient infrastructure requires a spectrum of interventions. For corridors with the dynamic and variable traffic flow typical of a university town like Morgantown, the AI-enhanced algorithm described in Part I offers a future of unparalleled efficiency. Yet, for countless rural highways where the immediate need is preventing tragedy, the accessible safety intervention of the intelligent warning beacon detailed in Part II provides an essential and immediate first step. Together, they form a comprehensive toolkit designed to meet the region's diverse needs.
This strategic approach, developed at West Virginia University by students who know these roads firsthand, is fundamentally Appalachian in its character. It acknowledges the region's reality—the need for practical, cost-effective safety measures on its rural arteries—while simultaneously championing the local innovation required to build a smarter future for its growing centers. Ultimately, this project is about more than just managing traffic; it is about creating an infrastructure that is responsive to its people, finally allowing the roads of the Mountain State not just to be seen, but to see.
References
U.S. Department of Energy, Alternative Fuels Data Center. (2021, May). Idle reduction for personal vehicles (Publication No. DOE/GO-102021-5586).
https://afdc.energy.gov/files/u/publication/idling_personal_vehicles.pdf
U.S. Department of Energy, & U.S. Environmental Protection Agency. (n.d.). Driving habits. FuelEconomy.gov. Retrieved October 3, 2025, from https://www.fueleconomy.gov/feg/driveHabits.jsp
U.S. Department of Transportation, Federal Highway Administration. (n.d.). About intersection safety. Retrieved October 3, 2025, from https://highways.dot.gov/safety/intersection-safety/about
Back to Directory