A groundbreaking study from MIT suggests that “eco-driving” techniques, which involve automatically controlling vehicle speeds to minimize stops and idling at intersections, could drastically reduce carbon emissions in urban areas. Detailed in the journal Transportation Research Part C: Emerging Technologies, this research used advanced AI to simulate traffic patterns at numerous intersections across three major U.S. cities.
The findings are promising, indicating that adopting eco-driving could reduce yearly carbon emissions at intersections by 11-22%. This method fine-tunes vehicle speeds to ease traffic flow without compromising traffic speed or safety.
In the short term, eco-driving could be facilitated through smartphone apps providing speed guidance. Looking ahead, such technologies might be integrated directly into semi-autonomous and fully autonomous vehicles.
Remarkably, even minimal adoption could yield significant benefits. If just 10% of vehicles participate in eco-driving, cities could achieve 25-50% of the potential CO2 reductions, as smooth traffic flow benefits even non-participating cars. Targeting just 20% of key intersections could capture 70% of the total emission reduction potential.
Senior author Cathy Wu emphasizes the practicality of eco-driving, noting that with smartphones already in cars and the rapid evolution of vehicle automation, implementing this simple, scalable solution could be just around the corner.






