Tesla to Build Own Maps to Fix FSD Speed Limits and Routing

Tesla plans to build its own map data to address problems with its Full Self-Driving (FSD) system. The company says this new layer will help the vehicle read speed limits, handle parking, and choose routes. The announcement came from Tesla’s head of AI during the third-quarter 2026 earnings call. This move marks a shift for the company. Its leadership has long argued that self-driving systems relying on maps are brittle. The new plan suggests the fleet will now map roads to help the system navigate.

The request came from a shareholder during the earnings call. The investor noted that mapping continues to be a problem for FSD. They cited speed limits, parking, and typical routes as the biggest issues. Tesla’s VP of AI, Ashok Elluswamy, responded to the concern. He said the company “shouldn’t be depending on external people’s maps.” Elluswamy explained that Tesla is building its own maps with a new AI-based architecture. This system will learn from the collective fleet and individual driver preferences. It will also respond to voice commands.

Tesla’s stance on mapping has changed over time. In 2015, the company built high-precision maps using data from customer cars. By 2019, Elon Musk called high-precision maps a “huge mistake.” He said relying on them makes a self-driving system “extremely brittle.” The company aimed to solve driving with cameras and AI alone. Now, seven years later, Tesla is building a map layer to fix things its vision system gets wrong. The company is using its own fleet to collect data for this purpose.

Tesla is not building centimeter-level HD maps. However, it is creating its own navigation data to support FSD. The issues Tesla aims to fix—speed limits, parking, and routes—relate to understanding the environment. FSD often fails at these tasks. In October 2026, a Belgian road safety group found that FSD sped in 55% of the 30 km/h zones tested in Brussels. The system averaged 44 km/h in those areas. In the US, NHTSA is weighing an investigation into claims that Tesla Robotaxis speed through active school zones. This affects about 1,000 Model Y and Cybercab vehicles.

This approach resembles how Waymo operates. Waymo maps a city with sensor-equipped vehicles before launching service. It keeps maps updated as its fleet drives. Tesla is now describing a similar process. Its cars collect data to feed a Tesla-built map. This helps every car drive better. The main difference is that Waymo’s cars use lidar. Tesla’s cars do not.

Tesla’s Robotaxi service operates city by city. It runs inside small geofenced areas. Tesla handles fleet operations in these zones. The company is not talking about HD maps today. However, a fleet-built map that grows more detailed could lead that way. Tesla has pushed unsupervised FSD in customer cars to the second quarter of 2027. This will happen only inside its Robotaxi fleet.