Tesla Adopts Its Own Maps to Fix 55% Speeding Issue in Brussels

Drivers who rely on Tesla’s Full Self-Driving system may soon see fewer sudden corrections when approaching intersections or parking lots. The company is currently developing its own mapping infrastructure to address frequent errors in reading speed limits, finding parking spaces, and selecting common routes. This development signals a practical change for daily commuters who have grown accustomed to the system occasionally misinterpreting simple road signs.

During the Q3 2026 earnings call, Ashok Elluswamy, Tesla’s VP of AI, confirmed that the company is moving away from third-party navigation data. He stated that Tesla should not depend on external providers for map data. Instead, the firm is constructing a new architecture that uses AI to build maps from the collective experience of its vehicle fleet. This system will also learn individual driver preferences and respond to voice commands, aiming to create a more personalized driving environment.

The rollout is planned to begin with the start and end of trips in the upcoming FSD v15 release. Over the following months, this internal map data is expected to take over navigation, parking, and destination handling. This shift addresses a long-standing point of contention among tech analysts and enthusiasts, who had noted that Tesla’s approach to self-driving had diverged significantly from competitors like Waymo, which relies heavily on high-definition maps and lidar.

This announcement marks a significant departure from the company’s public stance for the past decade. In 2019, CEO Elon Musk dismissed high-precision lane maps as a "huge mistake" and a "really bad idea" that made self-driving systems brittle. At the time, he argued that Tesla could solve driving with cameras and AI alone, without the need for the detailed map layers used by other autonomous vehicle developers. He even referred to lidar as a "fool's errand," setting a tone that Tesla’s vision-based approach was superior and more robust.

However, recent data suggests that the vision-only system still struggles with basic environmental understanding. Last month, a Belgian road safety group found that Tesla vehicles exceeded speed limits in 55% of the 30 km/h zones tested in Brussels. The average speed was 44 km/h, often because the car’s display showed a 50 km/h limit where one did not exist. In the United States, the National Highway Traffic Safety Administration is also weighing an investigation into claims that Tesla Robotaxis speed through active school zones, covering approximately 1,000 Model Y and Cybercab vehicles.

The decision to build its own maps brings Tesla’s strategy closer to that of Waymo. Both companies now use their fleets to collect data that feeds into a central map, helping every vehicle drive better. While Waymo’s vehicles are equipped with lidar, Tesla’s cars rely on cameras. The main difference lies in the sensor technology, but the underlying logic of using fleet data to create a shared knowledge base is now similar. Tesla is not currently building centimeter-level high-definition maps, but the system is designed to become more detailed over time.

This approach also aligns with how Tesla’s Robotaxi service operates. The service runs in specific geofenced areas within cities, where the company manages the fleet operations. By building a map that keeps getting more detailed, Tesla is effectively creating a localized knowledge base for these service areas. This is a pragmatic step that acknowledges the limitations of pure vision in complex urban environments, where speed limit signs and parking rules require precise, consistent data.

For consumers, this means a potential improvement in the reliability of the system they are already paying for. If the new map layer successfully corrects speed limit errors, it will make the driving experience safer and less stressful. It is a sign that the company is prioritizing real-world performance over its previous philosophical objections to using maps. The goal is to make the system work more like a human driver who knows the local roads, rather than a machine that has to guess every time.

There are still open questions about how quickly the new map data will be available to all users. The integration of individual driver preferences and voice commands adds a layer of complexity to the system. However, the core objective is clear: to reduce the frequency of errors in the most common driving scenarios. As Tesla continues to refine its AI, the line between a vision-based system and a map-based one is becoming increasingly blurred.

The shift also highlights the importance of data in the race to autonomous driving. Tesla has millions of cars on the road, which provides a vast amount of real-world data. By leveraging this fleet to build its own maps, the company is turning its customer base into a tool for improving the technology. This is a strategic advantage that other companies may find difficult to replicate, as they do not have the same scale of vehicle deployment.