Tesla Inc. (TSLA) is building an AI-based mapping system that will draw on information from its vehicle fleet and learn from both individual preferences and patterns across its cars.
That could make mapping more integral to autonomous driving. But before anyone declares war on Google Maps, a caveat: Tesla has not said it is directly trying to replace Alphabet Inc.'s (GOOGL) Google (GOOG) Maps as a consumer app.
"We shouldn't be depending on external people's maps," Ashok Elluswamy, VP of AI at Tesla, said during the company's third-quarter earnings call Q&A.
"So we are building our own maps," he added.
Super Scrutiny
Elluswamy said Tesla is developing "an entirely new architecture with essentially AI-based mapping and navigation." The company's vehicle fleet is central to that approach: "The collective intelligence of all the cars is going to flow into making the maps and navigation be super cool," he said.
The ambition is to build navigation around data Tesla can gather and update itself, rather than depending entirely on an outside mapping provider. That may matter for a company seeking to expand its driverless ride-hailing service, where navigation must work alongside the vehicle's ability to understand roads, reach destinations and handle the final stretch of a trip.
Tesla's transition to an AI-based mapping and navigation system also faces severe scrutiny as safety advocates and federal regulators warn that relying on a camera-only approach introduces critical visibility vulnerabilities. Last month, the National Highway Traffic Safety Administration escalated its investigations into the automaker's Full Self-Driving software, currently being used in Austin, Texas.
Beyond Turn-by-Turn Directions
Tesla says the system is expected to reach beyond route selection. Elluswamy said the technology would start with trip start and end points "pretty soon" as part of the V15 release, with navigation functions such as parking and destination handling expected to roll out over the following months.
He also said the system would learn "your personalized preferences plus also the fleet preferences."
That suggests Tesla wants navigation to improve based on both a driver's habits and fleet-wide information. The transcript doesn't spell out how the system will be built, what external map data it may still use, or how it will measure accuracy against established services.
The Google Maps Question
Google Maps plays a powerful role in everyday navigation, but Tesla's stated project is not evidence of a direct challenge to a consumer app. The more immediate implication is strategic: autonomous vehicles may need mapping and destination-handling tools designed around the vehicle's needs, not just the person choosing a route.
For investors, the test is execution. Watch for the V15 rollout, the pace at which Tesla replaces existing navigation functions, and evidence that the system improves parking and destination handling in real-world use. If those capabilities become important to Robotaxi operations, Tesla's mapping work could become a meaningful part of its autonomy stack, not merely a new feature in its cars.
