Improving NYC Congestion Through Smart Traffic Systems
Group 4: Adam Asad, Ayaan Rahman, Matt Leon, & Michael You
The City College of New York
ENGL 21007
Professor Julia Brown
April 22, 2025

 

Summary:
Traffic congestion is a long-standing issue in big cities like New York City, particularly in highly populated zones like Midtown Manhattan. It leads to increased commute times, greenhouse gas emissions, and reduced urban mobility and efficiency in general.
Our proposal envisions the installation of smart traffic lights with live cameras and artificial intelligence to self-adjust signal patterns based on actual traffic volume in real time. Unlike traditional fixed-time systems, these adaptive traffic lights can respond in real time to fluctuations in traffic flow, allocating more time to high-capacity lanes and breaking up the traffic jams. With the help of AI technology, this system guarantees to lower delays, traffic efficiency, and vehicle emissions across impacted intersections. Comparable technologies have been successfully implemented in other cities all over the United States, resulting in notable gains in traffic efficiency and environmental concerns.
Our team consists of engineering students who have conducted the proper research on smart traffic technologies and urban infrastructure systems as a part of this proposal. While we aren’t professionals yet, our project draws and utilizes existing case studies and engineering principles to present an innovative solution to New York City’s traffic mess.
The project is estimated to cost approximately $750,000 to meet system development, equipment installation, and pilot deployment in Hell’s Kitchen, Manhattan. We believe that this investment will be worthwhile in the long term both for city residents and planners in the form of a more efficient and sustainable transportation system.

Table of Contents

Summary: 1
Introduction 2
Project Description 4
Budget 7
Long Term Effects 9
Conclusion 10
References 11

 

 

 

Introduction
New York City suffers from severe traffic congestion especially in densely populated areas like midtown Manhattan. It ranks among the most congested cities in the United States. This congestion leads to delays and also contributes to increased carbon emissions due to idling vehicles. Traditional fixed time traffic signals are not able to respond to real time traffic patterns which leads to inefficient flow and blockage.
We proposed implementing AI powered traffic light systems throughout Manhattan. This system would use real time cameras, and GPS sensors to adjust light signals based on real time traffic conditions. The goal is to reduce congestion, improve traffic flow, lower carbon emissions, and overall increase the efficiency of transportation in the city.
This idea has already been tested in various cities like Pittsburgh, Boston, and Copenhagen. They have piloted and adopted these systems with notable success. For instance Boston partnered with Google’s Project Green Light team and worked together to use AI to model traffic patterns and create signal timing recommendations that can reduce traffic and emissions. Since implementing this, Boston saw a 50% reduction in stop-and-go traffic at key intersections. In the journal Information published by MDPI, researchers from the University of Alicante’s Department of Computer Engineering showed reductions in emissions, idle time, and travel delays when smart traffic lights were implemented.
The scope of this proposal focuses on describing the design and function of AI-powered traffic systems and evaluating their feasibility in New York City. We will analyze the benefits and potential challenges of this new system, and provide a detailed cost breakdown for the proposed pilot in Hell’s Kitchen. Lastly, we will be reviewing successful case studies from other cities to identify the long-term value and sustainability of this solution in NYC.

Project Description
Our proposed solution is to implement AI powered traffic lights in Manhattan to help reduce traffic congestion, emissions, and improve traffic flow. These smart traffic lights will use real time cameras and GPS sensors to adjust traffic signals. Unlike outdated fixed time traffic lights these smart lights have the ability to adjust dynamically based on actual traffic conditions.

Feasibility
This solution is very feasible as similar AI powered traffic lights have already been piloted or been in use. For example, In Pittsburgh, an AI system reduced travel time by 25% and cut vehicle idling time by 40% (Smart Cities Dive). In Denmark, SWARCO’s smart traffic management system uses GPS and camera data to optimize traffic lights, costing between €20,000 – €40,000 per intersection. This cost is comparable to other traffic system upgrades like replacing signal heads or installing sensors, which makes it financially feasible for a pilot and eventual expansion.
We can assess the feasibility of this system in NYC by analyzing traffic volume data and congestion patterns from NYC Open Data and NYPD TrafficStats. We can also perform technical reviews of intersections in NYC to determine infrastructure readiness for smart sensors and AI integration. Lastly, we will consult with NYC DOT, and urban planners.

Benefits and Consequences
Implementing AI powered traffic lights in NYC would bring many benefits. First, it would significantly reduce traffic congestion, which currently costs the city an estimated $11 billion annually in lost productivity and fuel waste (INRIX Inc., 2019). That is roughly about $2,000 being lost per driver. By upgrading signal timings in real time, the system can not only help reduce average commute times but also reduce vehicle idling time, contributing to greenhouse gas emissions. Additionally, smoother traffic flow benefits emergency vehicles by clearing intersections more efficiently, especially with roads that don’t have service lanes.
If we do not decide to implement these new smart traffic lights it would lead to severe consequences. For example, with the old traffic light system and its fixed timings, it causes avoidable stops even when the roads are clear. Leading to unnecessary wasted time and fuel which directly leads to lost money. With more idling time there is also more emissions that release into the atmosphere which hurts our planet and our health.
Graph Example:
The graph demonstrates the impact of AI traffic management in Pittsburgh and Boston(two big cities in the same region as NYC) by comparing significant traffic statistics before and after adoption. In Pittsburgh, travel time decreased by 25%, idling time by 40% (Smart Cities Dive), and CO₂ emissions decreased by 21% in the SURTRAC system (IEEE Spectrum, 2024). Google-sponsored Project Green Light in Boston reduced stop-and-go crashes by 50% (Cities Today, 2023).

Methodology
Phase 1: Pilot program
First select 7-10 high traffic interactions in Manhattan. Next, Install AI powered traffic light systems with cameras and sensors. We will then analyze the data to monitor its performance.
Phase 2: Evaluation
Compare the performance of the new smart traffic light system to the old system. Use metrics like average wait time, vehicle emissions, and accident frequency to evaluate effectiveness.
Phase 3: Roll Out
If the pilot program was successful, roll out the new system in additional areas of NYC.
Resources Needed
Funding of approximately $750,000 for pilot phase.
Partnerships with vendors (e.g., SWARCO, GoodVision).
Technical team to handle system installation and data analysis.
Obstacles and Challenges
A potential challenge of this project is privacy. Many people might worry about being surveilled by real-time traffic cameras. To address this, the cameras and sensors only track vehicle movement and traffic density, not faces or personal identities. Transparency and clear data policies would also be implemented to reassure the public.

Budget
To effectively implement an AI powered adaptive traffic signal system in Midtown Manhattan, we propose a targeted pilot program in Hell’s Kitchen, a neighborhood known for its consistently high traffic volume and congestion issues. Before deploying citywide, focusing on this smaller region allows for a more accurate assessment of impact and feasibility. For instance, systems like the one implemented by SWARCO in Denmark cost between €20,000 and €40,000 per intersection, depending on location and complexity (SWARCO, 2020). This covers the necessary hardware, including real time cameras, adaptive control units, and signal updates. The pilot would focus on 20 key intersections throughout Hell’s Kitchen, particularly along 9th and 10th Avenues, nearby feeder streets where congestion is most severe during peak hours, (NYC Gov, 2014), the base hardware installation would amount to $400,000.

In addition to the hardware, the success of this AI powered traffic system relies on having a strong software infrastructure in place. This includes advanced AI platforms that process traffic data in real time, adjusting signal timings based on the flow of traffic. The costs for software licensing and integration are projected to be between $100,000 and $150,000, reflecting the complexity and scope of such a system (SWARCO, 2020). These one-time costs include an initial licensing fee as well as annual renewal charges, which ensure continued access to updates and technical support. Several cities, including Pittsburgh and Boston, have already seen success with similar technologies. For example, Boston’s Project Green Light used AI and Google data to reduce stop and go traffic by 50%, showing how well this kind of system can improve traffic flow and reduce congestion (Boston Gov, 2021). The software costs will cover things like machine learning algorithms, cloud based data analytics, and integration with New York City’s existing traffic control systems. These technologies rely on live video feeds and cloud processing to optimize signal timing in real time, as seen in systems like those used by SWARCO and Parquery (Parquery, 2022).
To ensure the system continues to operate at peak efficiency, we would also need a two year service contract. This would cover regular updates, technical support, and essential staff training. The service contract is expected to cost between $50,000 and $75,000. Training is particularly important for Department of Transportation technicians and NYPD traffic controllers, who will be responsible for managing the system independently after the two-year period ends. Once the period ends, a new budget for the workforce and their return should be discussed based on the success of the project. Ensuring that they are fully trained and supported will be key to the long term success of the project, helping to avoid technical issues and guaranteeing that the system remains adaptable to future traffic challenges.
The overall cost for the pilot implementation in Hell’s Kitchen is projected to be between $600,000 and $750,000. While this may appear substantial, the long term benefits including reduced congestion, improved traffic safety, and lowered carbon emissions represent a strong return on investment.
Long Term Effects
While our original plan is to have a pilot program in Hell’s Kitchen, our long-term goal is to create a model that can be implemented throughout all five boroughs of NYC. Information collected in the pilot will not only enable us to track short-term effects, but also will inform citywide planning. Over time, the system can be extended to Queens’ high-capacity corridors, Brooklyn, the Bronx, and Staten Island in a bid to extend the traffic system to be fully integrated and responsive.
As the city grows, so will the demand for smarter mobility solutions. AI traffic systems have an upgradeable system that can become embedded in newer technology, such as integration with autonomous cars, dynamic public transport priority, and even climatically responsive traffic planning. AI platforms are engineered in a manner that the system will not be outdated, but rather will be in a position to evolve for decades.
This vision aligns as well with NYC’s long-range goals of reducing carbon emissions, improving public health, and building more livable communities. With planning for scale ahead of us, we are positioning ourselves toward a more thoughtful, better city that is capable of responding to the needs of New Yorkers over the years.
Conclusion
NYC does not have to be stuck with outdated traffic systems while the rest of cities move forward. AI controlled traffic lights are no concept for tomorrow, as they’re already in cities like Pittsburgh and Boston. The systems are speeding up traffic, reducing pollution, and making people’s daily commute less painful. This is not reinventing the wheel. This is implementing solutions that work in one of the world’s busiest cities.
If we stick with the same old traffic system, the issues we have with traffic congestion, wasting time, and pollution will only increase. But by making the change, even with a tiny pilot program, we can start to make some real changes in how people travel around our city. This is an opportunity for New York to show that NYC is ready to embrace intelligent and sustainable solutions that make life better. AI traffic systems will not solve everything, but they can improve the everyday life of NYC residents significantly. Let us take that step to making our streets smart and working better for everyone.

 

References
Boston’s Mayor’s Office. (2024, August 8). City of Boston Partners with Google on Traffic Signal Optimization. Boston.gov. Retrieved April 22, 2025, from https://www.boston.gov/news/city-boston-partners-google-traffic-signal-optimization
Boston’s Mayor’s Office. (2024, August 8). City of Boston Partners with Google on Traffic Signal Optimization. Boston.gov. Retrieved April 22, 2025, from https://www.boston.gov/news/city-boston-partners-google-traffic-signal-optimization
INRIX Inc. (2020, March 9). INRIX: Congestion Costs Each American Nearly 100 hours, $1400 A Year. INRIX. Retrieved May 6, 2025, from https://inrix.com/press-releases/2019-traffic-scorecard-us/
nyc.gov. (2014). Clinton/Hell’s Kitchen Neighborhood Traffic Study. nyc.gov. https://www.nyc.gov/html/dot/downloads/pdf/2014-hells-kitchen-final-report.pdf
Parquery. (n.d.). Green on Demand. Control Traffic Lights with Cameras not Induction Loops. https://parquery.com/control-traffic-signals-with-cameras-not-induction-loops/
Snow, J. (2017, July 20). This AI Traffic System in Pittsburgh has Reduced Travel Time by 25%. smartcitiesdive.com. https://www.smartcitiesdive.com/news/this-ai-traffic-system-in-pittsburgh-has-reduced-travel-time-by-25/447494/
SWARCO. (n.d.). AI-based Traffic Management. SWARCO. Retrieved April 22, 2025, from https://www.swarco.com/stories/ai-based-traffic-management
IEEE Spectrum. (2024, April 12). Pittsburgh’s smart traffic signals will make driving less boring. IEEE Spectrum. Retrieved May 6, 2025, from https://spectrum.ieee.org/pittsburgh-smart-traffic-signals-will-make-driving-less-boring
Cities Today. (2023, March 22). Boston uses AI to reduce stop-go traffic by 50 percent. Cities Today. Retrieved May 6, 2025, from https://cities-today.com/boston-uses-ai-to-reduce-stop-go-traffic-by-50-percent/