Google Study Finds Rerouting Under 2% of Trips Lifts Speeds 2% and Cuts CO2
Updated
Updated · Google Research · Jul 7
Google Study Finds Rerouting Under 2% of Trips Lifts Speeds 2% and Cuts CO2
1 articles · Updated · Google Research · Jul 7
Summary
10 major U.S. cities showed measurable gains when Google Maps altered routes for fewer than 2% of observed trips, according to a Google Research study published in Nature Cities.
A six-month experiment steered drivers away from about 100 pre-selected bottleneck segments per city toward similar-time alternatives, using treatment and control days to isolate the effect.
Targeted segments saw a median 2% increase in driving speeds, while fuel consumption rates fell 0.5% to 1.0%; across all affected roads, speeds rose about 0.35% overall and 0.5% at peak hours.
Google said dispersing traffic from major bottlenecks onto a wider set of peripheral roads produced network-wide benefits for app users and non-users alike, with potential savings of thousands of tons of CO2e per city each year.
The study positions network-aware routing as a template for broader smart-city traffic management, including real-time optimization and coordination with connected vehicles and infrastructure.
As algorithms reroute traffic for city-wide benefit, who decides the acceptable trade-off for individual drivers sent on longer routes?
With Google proving its model, can national ITS programs compete or must they collaborate with big tech to succeed?
Could centrally managing traffic flow create new vulnerabilities, making cities susceptible to algorithmic errors or cyberattacks?
Google's AI Rerouting Cuts Urban CO2 by Up to 1%: A 10-City Study on Smarter Traffic and Cleaner Air
Overview
In July 2026, Google published a groundbreaking study showing that subtle, data-driven changes in urban traffic can make cities more efficient and improve air quality. Over six months, Google ran a large-scale experiment in 10 major U.S. cities, where less than 2% of trips were rerouted. By carefully selecting about 100 road segments in each city based on past congestion, the system made small navigation adjustments to spread out traffic and ease pressure on busy areas. This intelligent rerouting improved traffic speeds and demonstrated that smart, targeted interventions can optimize urban mobility without expensive infrastructure changes.