Strava updates its AI route planning and cheater detection

Strava is making it easier to plan workouts and keep activity leaderboard rankings fair. The updates rolling out over the coming weeks focus on helping users optimize their workout routes to compete against other users and their own personal bests, building on some of the existing AI features that Strava announced last year.

Anyone paying for a Strava subscription (starting at $11.99 monthly) can now access a new AI-powered routes experience under the Maps tab that should provide more intuitive suggestions based on popular routes enjoyed by other Strava users. Users can generate community-backed routes from custom starting points or their current location, pulling data from Strava’s heatmaps feature.

Other route-related updates will be rolled out to the Strava mobile app in the coming months, including changes to the tappable points of interest (POI) feature that currently enables subscribers to instantly generate routes to cafés, restrooms, viewpoints, and other locations. Starting in June, POI’s will also display elevation, distance, and estimated arrival time information, and allow users to upload photos of the location. Point-to-point routing will also launch in July, which uses heatmaps and machine learning to deliver “the most efficient, activity-specific route from A to B,” according to Strava.

Strava is also doubling the number of live segments, which allow users to view real-time performance data and achievements in sections of their route, and introducing additional data screens for subscribers.

Finally, Strava says it’s “continuing to advance” the AI-powered Leaderboard Integrity feature it launched to weed out cheaters on cycling and running paths. The company says that 4.45 million activity logs have been removed so far that carried the wrong sport type, or were recorded in vehicles — which is an easy way to fabricate scores now that e-bikes can make anyone the king of a mountain.