Urban traffic congestion is one of the world’s most persistent and dangerous challenges, contributing to an estimated 1.19 million annual fatalities and countless disabilities. Beyond safety concerns, it fuels major inefficiencies and environmental damage through long waiting times and increased CO₂ emissions. The ripple effects of traffic extend well beyond the roads themselves wasted hours in gridlock diminish productivity, delayed emergency vehicles can cost lives, and unchecked emissions accelerate climate change. For rapidly growing cities across the globe, the question is no longer whether traffic management is important, but how innovation can radically reshape it.
Against this backdrop, Aarav Brahmbhatt, a rising researcher in the YRI Fellowship, set out to address this global issue with cutting-edge artificial intelligence. During his time in the Fellowship, Aarav designed a system that goes beyond traditional traffic detection. Instead of simply monitoring congestion, his framework uses AI to predict accident likelihood, waiting times, and environmental impact giving city planners a scalable, low-cost tool for intelligent traffic management that could prevent tragedies before they unfold.
At the core of his research, Aarav trained a YOLO-based object detection model, a highly efficient neural network architecture designed to identify and classify objects in real time. By analyzing aerial and CCTV traffic images, his model could accurately spot and categorize vehicles from motorcycles weaving through crowded lanes to large trucks dominating intersections. Once this raw visual data was processed, Aarav paired the extracted features with Random Forest regressors, a machine learning technique renowned for its robustness and interpretability. Together, these tools allowed him to forecast key metrics such as accident probability, congestion-induced waiting times, and even CO₂ emissions from idling vehicles.
The result is a high-accuracy predictive model capable of not just observing traffic but anticipating its risks and consequences. This marks a fundamental shift in how traffic management systems can operate. Instead of reacting to accidents after they occur, cities could be alerted to potential hotspots in advance, allowing preventive measures such as re-routing traffic, adjusting signal timings, or deploying road safety patrols.
“I wanted to create something that not only detects traffic, but also predicts future problems,” Aarav explained. “If we can foresee traffic concerns before they get out of control, we can intervene early, save time, reduce emissions, and avoid accidents.” His words underscore a vision of traffic systems that are proactive rather than reactive an evolution that could save thousands of lives each year.
This visionary project reflects what sets the YRI Fellowship apart: empowering ambitious students to tackle real-world problems with advanced research and mentorship. Unlike conventional programs that limit high school research to classroom experiments, the Fellowship challenges students to engage with issues of global significance, pairing them with leading mentors who can guide them through technical, ethical, and practical dimensions of their work. Aarav’s progress illustrates the power of that model: given the right platform, even young innovators can create solutions with international relevance.
Urban planners and policymakers worldwide have long sought effective tools to manage traffic chaos. Traditional solutions like expanding road networks, adding flyovers, or increasing public transit fleets often take years and billions of dollars to implement, with mixed results. Aarav’s AI-driven approach offers a scalable, cost-efficient alternative that could be deployed with existing surveillance infrastructure. Cities already equipped with CCTV and aerial imaging could adopt predictive modeling with minimal additional investment, making it especially attractive for developing regions where budgets are tight but road safety is a pressing concern.
The environmental dimension of Aarav’s work also carries profound implications. According to studies, vehicles idling in congested traffic contribute disproportionately to urban air pollution, releasing harmful greenhouse gases and particulate matter. By anticipating where congestion is likely to occur and implementing early interventions, cities could reduce unnecessary emissions, improving both public health and climate resilience. Aarav’s integration of CO₂ emission forecasts into his model ensures that traffic safety and sustainability are addressed together, rather than as separate policy goals.
Beyond the technology itself, Aarav’s journey demonstrates the growing role of youth-led innovation in solving complex global problems. With access to powerful computing tools, open-source models, and mentorship networks like YRI, today’s students are no longer limited by age or geography. Their contributions can shape smarter, safer, and greener cities, challenging the notion that only governments or large corporations hold the keys to transformative solutions.
By combining object detection, predictive modeling, and environmental analysis, Aarav’s project shows how the next generation of scientists is driving change at the intersection of technology, public safety, and sustainability. His work is not just a technical achievement but also a call to action reminding us that innovation thrives when young minds are empowered to experiment, iterate, and imagine boldly.
The story of Aarav and his research is a testament to what happens when curiosity meets opportunity. It signals a hopeful future where cities worldwide may one day predict and prevent accidents, minimize congestion, and curb emissions all powered by AI solutions designed by visionaries who are still in high school.
Learn more about the Fellowship and how students like Aarav are transforming the future of science at yriscience.com.
