Tesla says FSD's next maps will come from what its customers' cars see, not from survey crews. That matters to anyone who has watched the system misread a speed limit or pick an odd route. We covered the reversal in our earlier story on Tesla and Waymo. This piece is about how the thing is meant to work.
What Tesla says it is building
According to Gadget Review's account of the Q3 2026 earnings call, a retail investor asked why Tesla leans on third-party maps when speed limits, parking and typical routes remain FSD's biggest problems. The call transcript published by the Motley Fool carries the same question. Yahoo Autos quotes VP of AI Ashok Elluswamy agreeing: "we shouldn't be depending on external people's maps". In the same report he described "an entirely new architecture with essentially AI-based mapping and navigation".
Going by his words on the call, the design has three parts. It absorbs what the whole fleet learns, which he called "the collective intelligence of all the cars". It learns each driver's preferences, a point Drive Tesla Canada also quotes. And it accepts voice requests for different routes. Yahoo Autos reports that Tesla currently relies on Google for map data, and Not a Tesla App likewise lists Google services for map tiles, imagery, points of interest and traffic.
Improvement every day is a promise
Nothing in the sources says how often the maps would update or how Tesla would vet fleet data before trusting it. Teslarati's coverage of the call lists map precision, validation, coverage, update frequency and privacy controls as unspecified. Not a Tesla App also notes that the fleet data collection methods were not detailed. So "better every day" is the logic of fleet learning, not a measured result.
The schedule is also a plan. Elluswamy said on the call that it begins with the start and end of trips in FSD v15, then extends to navigation, parking and destination handling over the following months. Gadget Review reports the same, and neither it nor Drive Tesla Canada gives a v15 release date. Treat that as a target.
How this differs from the old HD map
Gadget Review recalls that in 2019 Elon Musk called high-precision lane maps "a huge mistake" that leave self-driving "extremely brittle". Its write-up says the new layer isn't an HD map of the kind he criticized, and that Tesla has always used navigation maps. Waymo, by contrast, maps geofenced areas with dedicated lidar vehicles, while Tesla would draw on a far bigger consumer fleet, Teslarati notes.
The real test is whether known failures, such as misread speed limits, shrink. No independent data yet shows maps fixing that, and Teslarati's coverage points to no such figures.
Watch for v15's release notes and any outside speed-limit testing. Would you let your car's route and parking be shaped by what other owners' cars saw?
Sources
- 1.Tesla Is Building Maps for FSD After Years of Dismissal · Gadget Review
- 2.Tesla Is Building Maps for FSD After Years of Dismissal · Yahoo Autos
- 3.Tesla is building its own maps to fix Navigation issues · TESLARATI
- 5.Tesla (TSLA) Q3 2026 Earnings Call Transcript · The Motley Fool
- 6.Tesla Is Building Its Own AI-Based Maps, Will Begin Rolling Out In the Next Few Months With FSD v15 · Not a Tesla App
- 7.Tesla Building Own AI Maps and Navigation for FSD · Drive Tesla Canada
Reported by the WattsUpNext desk from the sources linked below. Spot an error? Tell us at corrections@wattsupnext.com.
The WattsUpNext Brief
Get stories like this in your inbox.
One email each weekday, only the topics you choose. Real news, sourced, no fluff. Unsubscribe in one click.
