Smarter Streets, Faster Progress for Autonomous Mobility?

MIT research suggests connected traffic systems could cut urban emissions by up to 22%, pointing to a faster, infrastructure-led path for cities

25 Jun 2025

At a busy city junction the greatest waste is not idling engines but impatience. Cars rush towards red lights, brake hard, then surge again. According to new research from the Massachusetts Institute of Technology (MIT), smoothing that simple behaviour could yield large rewards, well before fully driverless cars become common.

The study looks not at autonomy in the grand sense, but at eco-driving and intelligent speed control. Vehicles that can anticipate traffic signals slow gradually when a light is about to turn red, rather than racing and stopping. Using simulations of urban traffic, MIT’s researchers find that such foresight cuts stop-start driving, reduces fuel use and lowers emissions. In dense city settings, carbon output could fall by as much as 22%.

This is not evidence of a revolution already under way. The work models what connected traffic systems might achieve, not what they do today. But it points to a more modest and perhaps more realistic path for autonomous technology. Instead of waiting for cars that can handle every edge case, cities and firms might extract value from making vehicles better at reading, and obeying, traffic flow.

For the autonomous-vehicle industry, that is an appealing prospect. Technologies that allow cars to anticipate signals or congestion could work in partially automated vehicles, or even in human-driven ones. The prize is near-term gains: cleaner air, smoother journeys and less wasted time. Full autonomy can wait.

Some firms appear better placed than others. Mobileye specialises in mapping and prediction; Waymo has experience navigating crowded urban streets; Tesla commands a vast connected fleet. All could, in theory, benefit if cities adopt smarter signals and shared data. Yet there is no public evidence that any has deployed the specific eco-driving systems imagined by MIT. For now, alignment is conceptual, not commercial.

Infrastructure, long the neglected cousin of vehicle technology, is moving centre stage. Carmakers and mobility providers increasingly talk of working with cities. Smart signals, connected roads and data-sharing are pitched as tools to improve efficiency as much as sustainability.

The obstacles are familiar. Many American cities run on ageing traffic systems. The benefits of coordination rise with adoption, meaning early gains may be modest. Mixed traffic, some connected, some not, dilutes the effect.

Still, the message is clear. Autonomy is no longer only about removing the driver. It is about rethinking how traffic behaves. If cities learn to choreograph their streets more intelligently, cleaner and calmer travel may arrive sooner than the self-driving car.