Tesla’s head of AI, Ashok Elluswamy, told investors on the company’s third-quarter earnings call on Oct. 21 that Tesla is building its own maps for Full Self-Driving, or FSD, rather than leaning on outside map providers. A retail shareholder had asked why Tesla depends on third-party maps, saying the weakness shows up in speed limits, parking and the routes drivers usually take. Elluswamy answered, in substance, that Tesla should not be dependent on other companies’ maps.
Save more, waste less: The Cool Down’s newsletter sends smart, simple changes that help you and the planet in your everyday life.
We may earn a commission if you sign up.
The answer reverses a position Tesla’s chief executive took publicly seven years ago, and it arrives while regulators and safety groups are examining how FSD handles posted speeds. All of Tesla’s statements below come from Electrek’s coverage of the call, because Tesla’s own transcript and shareholder letter could not be opened for this report.
What Elluswamy said Tesla is building
According to Electrek’s Oct. 22 report, the plan is a new AI-based mapping and navigation architecture fed by learning across the whole Tesla fleet. The system is also meant to learn individual drivers’ preferences and to respond to voice commands. Elluswamy tied the timeline to software releases: starting with FSD version 15, the system will handle the start and end of trips, and over the following months it is expected to take over navigation, parking and destination handling.
The shareholder’s question set the terms. Mapping, the shareholder said, continues to be a problem for FSD in the three areas the question named, and Tesla’s reply was an architecture plan, not a list of bugs fixed or a date by which speed-limit errors would stop. Electrek’s report adds that the plan is also meant to learn from voice commands, so the system is pitched as a driver-facing navigation product as much as a fix for sign-reading.
Tesla’s earlier position on maps
Electrek traces how far the company has moved. In 2015, at the launch of Autopilot, Tesla told Fortune it was building high-precision maps with 100 times the granularity of standard navigation, using data from customer cars. At Autonomy Day in 2019, Elon Musk said Tesla had tried high-precision lane maps and called it “a huge mistake”, describing reliance on them as a really bad idea that makes a self-driving system extremely brittle. In the same presentation he dismissed lidar as a fool’s errand.
Electrek’s own view, offered as commentary rather than as a Tesla claim, is that using maps is not inherently contradictory and could be an advantage, but that Tesla’s approach now looks more like Waymo’s fleet-mapping method, with lidar as the main difference.
Electrek goes further in its own argument, saying Tesla’s driverless effort now relies on geofenced, map-based operation of the kind Musk once said did not count as true self-driving. The report notes that Tesla’s Robotaxi service runs city by city inside small geofenced areas. Whether those service areas and the new consumer-car mapping system end up sharing data is not something Tesla addressed in the answer Electrek summarized, and Electrek does not claim that they will.
The speed-limit evidence
The mapping question lands on documented problems. In Brussels, the pedestrian and cyclist safety nonprofit Johanna.be ran a study called FSD30, with testers driving about 400 kilometers over three days in July. Electrek reported on Sept. 24 that FSD in supervised mode exceeded the limit in 55% of the 30 km/h zones tested, averaging 44 km/h, and that the car’s screen showed a 50 km/h limit in most of the 30 km/h zones it drove through. The group’s project page is the link Electrek gives for the work.
The same kind of error is under scrutiny in the United States. Electrek says a defect petition, which reached NHTSA on Sept. 23 and was opened as file DP26010 on Oct. 9, alleges that Tesla Robotaxis speed through active school zones. NHTSA’s opening resume puts the affected fleet at about 1,000 Model Y and Cybercab vehicles, an estimate Electrek says may reflect vehicles certified in Texas rather than cars in service. Electrek’s coverage of the petition reports no crashes, injuries or deaths, and NHTSA has not decided whether to open a formal investigation. The petition cites three social media videos and four incidents in the agency’s own tally. Under federal rules NHTSA has 120 days from receipt to grant or deny the petition, which Electrek puts at around Jan. 21, 2027, and a grant would lead to a formal defect investigation.
What the maps plan does not settle
Tesla has not said that its maps will correct the Belgian pattern, and neither Elluswamy’s answer nor Electrek’s account links the speed-limit errors to specific map data. A car can misread a limit because of a sign, a database entry or a decision about which to trust, and a new navigation stack addresses only some of those. For owners who paid for FSD, the concrete commitments reported so far are the version 15 start and the months-long handover of navigation, parking and destination handling that Elluswamy described on Oct. 21.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
Save more, waste less: Subscribe to The Cool Down →
We may earn a commission if you sign up.
More from Morning Overview
- Two passengers died at an Ohio toll plaza, and the NTSB now wants the cash lanes shut
- Long use of a common prostate pill is tied to a higher chance of glaucoma
- Ford is recalling 223,472 F-150 pickups because the fuel tank can leak or detach
- Hybrids have 15% fewer problems than gas cars, while EVs and plug-in hybrids have about 80% more, Consumer Reports finds
