-
Introduction
-
Schematic Representation
-
Quantum Approximate Optimization Algorithm (QAOA) Vs. Classical Algorithm
Inspiration
The trip to Bitcamp 2024 at UMD, which should have taken only 40 minutes, ended up taking over an hour due to traffic. This frustrating experience sparked an intriguing idea, what if we could enhance the efficiency of traffic intersections? By reducing idling times and prioritizing traffic flow based on intersection density, we could significantly decrease pollution. However, controlling traffic on such a large scale presents a formidable challenge. Traditional computers struggle to handle the vast amount of data and to make decisions quickly and efficiently. This limitation led us to explore the potential of Quantum Computing. Utilizing the Quantum Approximate Optimization Algorithm, we were able to make decisions faster, more efficiently, and on a much larger scale.
What it does
Current traffic management systems are struggling to fulfill the Sustainable Development Goal 11, which aims to make cities inclusive, safe, resilient, and sustainable. Traditional traffic signal systems contribute to heavy and inefficient flow of traffic, increasing issues of congestion, particularly at key intersections. This not only causes significant delays and driver frustration but also increases fuel consumption due to prolonged idling and repetitive stop-and-go cycles, leading to higher emissions and greater environmental impact. Furthermore, congested roads heighten the risk of accidents. By implementing more efficient traffic management strategies, cities can enhance road safety, reduce congestion, and lower emission levels, aligning more closely with the objectives of SDG 11.
How we built it
Using Quantum Approximate Optimization Algorithm (QAOA), which is optimized for solving complex combinatorial optimization problems like traffic light configuration, to find the traffic signal with the heaviest traffic flow based on the graph. we also used a hybrid quantum-classical approach where a classical optimizer adjusts quantum gate parameters to minimize the cost function, refining these through iterative feedback from quantum measurements.
Libraries used: pennylane, numpy, matplotlib, cirq, NetworkX.
Challenges we ran into
Our biggest challenge was that we had limited knowledge prior to the last 15 hours. We had to attend many workshops here and study research papers to understand Quantum Computations and Algorithms from scratch. We had to make visualizations for traffic intersections at different scenarios and integrate various factors like prioritizing traffic flow based on traffic density among other things were few of the other challenges we ran into.
Accomplishments that we're proud of
Quantum computers offer a significant speed advantage over classical computers, enabling the processing of large datasets at unprecedented rates. This could potentially unlock our capability to allow for a real-time traffic management, which can greatly enhance the efficiency of traffic flow, reduce congestion, and subsequently lower fuel consumption and emissions. Additionally, we trained our model to reduce idling times contribute to decreased environmental impact which was our first priority - Sustainability. Our project directly supports Sustainable Development Goal 11, aiming to make cities and human settlements more inclusive, safe, resilient, and sustainable.
What we learned
We learnt to utilize quantum superposition which allow these computers to evaluate multiple traffic signals simultaneously, significantly enhancing their parallel processing capabilities. In addition, we learnt and used Specialized quantum algorithms, such as the Quantum Approximate Optimization Algorithm (QAOA), which are particularly well-suited for solving optimization problems and are used to iteratively refine solutions to reduce traffic congestion effectively. We have also learnt to use pennylane, which consists various libraries and used it in the project.
What's next for Traffic Optimization using Quantum Computation
The future of traffic optimization using quantum computation looks promising, focusing primarily on optimizing traffic light timings to ensure the most efficient and fastest routes for drivers. As quantum technology continues to advance, its capacity to manage larger and more complex systems will grow, providing a scalable solution that can adapt to future urban planning challenges. Currently, this approach assumes an ideal scenario, but we expect to introduce additional factors and parameters to fine-tune the model further. This expansion aims to enhance its capabilities and functionality, enabling it to represent and efficiently manage a large network of urban traffic.
Log in or sign up for Devpost to join the conversation.