Tesla Shifts Course, Building Its Own Maps for Self-Driving
The automaker is now developing proprietary mapping technology to address recurring issues with its automated driving features, a move that diverges from previous statements by its leadership.
By William Parker ·
Tesla has announced plans to develop its own mapping systems to resolve persistent difficulties with its Full Self-Driving software, including the accurate reading of speed limits, parking maneuvers, and route selection.
The disclosure regarding the creation of a proprietary mapping and navigation architecture was made by Tesla's Vice President of AI, Ashok Elluswamy, during the company's third-quarter 2026 earnings call. The initiative is designed to utilize data collected collectively by the vehicle fleet to generate map information, while also adapting to individual driver preferences and voice commands.
This development marks a distinct departure from the company's historical stance on the technology. In 2019, CEO Elon Musk characterized the reliance on high-precision lane maps as a huge mistake that renders self-driving systems extremely brittle. While Tesla previously utilized third-party navigation data, the decision to construct a proprietary map layer is intended to address specific failures in environmental understanding, particularly concerning speed limit signs and parking locations.
The company has previously experimented with high-precision mapping. In 2015, following the launch of Autopilot, Tesla indicated that it had begun building maps with a level of granularity significantly higher than standard navigation systems using data from customer vehicles. However, the firm later moved away from that approach, with Musk stating in 2019 that relying on such maps and lidar was a bad idea for a vision-based system.
Current data suggests that vision alone has not been sufficient to reliably handle certain driving tasks. A road safety group in Belgium recently reported that Full Self-Driving exceeded speed limits in 55 percent of tested 30 km/h zones in Brussels, with the vehicle's screen displaying a 50 km/h limit in most instances. Additionally, the National Highway Traffic Safety Administration is considering an investigation into reports of Tesla vehicles speeding through active school zones.
Tesla's approach to mapping now closely mirrors the methodology employed by competitors such as Waymo, which utilizes sensor-equipped vehicles to map cities prior to service launch. While Waymo relies on lidar for this purpose, Tesla intends to use data collected by its fleet to construct a map that assists all vehicles in driving more effectively. The company's Robotaxi service currently operates within specific geofenced areas, a strategy that contradicts earlier assertions that the technology would function as a general solution anywhere through vision alone.
The rollout of the new mapping technology is scheduled to begin with the handling of trip start and end points in Full Self-Driving version 15. Over the subsequent months, the system is expected to expand its capabilities to include navigation, parking, and destination handling.
Source: Electrek
