Stop-and-go waves on freeways are not just annoying. They are a direct emissions source. Smoothing them — eliminating the accordion pattern where cars repeatedly brake and accelerate — cuts CO₂ by 8-12% and CO by 14-29%.
The study (arXiv:2603.21476) reconstructed smooth, feasible trajectories from nine weeks of real data on I-24 MOTION, a heavily instrumented testbed in Nashville. The method: quadratic programming to find the smoothest path each vehicle could have taken while maintaining a maximum gap of 0.1 miles from the car ahead. Same origin, same destination, same lane, same boundary conditions. Just without the waves.
The critical number is the gap constraint. Previous work on wave smoothing assumed unrealistically large following distances, making the benefits theoretically impressive but practically useless. At 0.1 miles — close to normal following behavior — the CO₂ savings are still 8-12%. The NOₓ reduction reaches 31%.
The mechanism is engine load variance. In smooth flow, engines operate near their efficient cruise point. In stop-and-go, they cycle between idle and peak acceleration — the least efficient regime. The emissions aren't proportional to average speed; they're proportional to the variance in acceleration. Two vehicles traveling the same average speed produce wildly different emissions depending on whether that speed was constant or oscillating.
This inverts the usual framing. The problem isn't that cars move — it's that they move unevenly. The pollution lives in the derivative, not the function. The smoothest commute isn't the shortest one. It's the steadiest.