Tesla has made another step in the Full Self Driving (FSD) capability of cars.
Technologies Used
Tesla Vision (Camera-Based Perception)
Tesla's official support page details their camera-based system, stating that all vehicles built for the North American market now use their camera-based Tesla Vision to deliver Autopilot features, rather than radar.
Tesla AI & Neural Networks
Tesla's AI & Robotics page discusses their development of AI training chips to power their Dojo system, which is integral to training their neural networks for self-driving capabilities.
Full Self-Driving (FSD) Software
Tesla's support page on Full Self-Driving (Supervised) Subscriptions provides insights into the features and capabilities of their FSD software, emphasizing that it requires active driver supervision.
Tesla Hardware (HW) Platforms
The Wikipedia page on Tesla Dojo provides an overview of Tesla's custom-built AI chips and the architecture of their Dojo supercomputer, which is designed to enhance their self-driving technology.
Path Planning & Decision-Making
Tesla's official Autopilot page explains how their system enables the car to steer, accelerate, and brake automatically within its lane, highlighting the decision-making processes involved.
High-Definition Maps (Limited Use)
While not heavily detailed in the provided sources, Tesla's approach focuses on real-time scene understanding through their vision-based system, as outlined in their support documentation.
Tesla Fleet Learning
Tesla's Vehicle Safety Report emphasizes the role of data collected from their global fleet in enhancing the safety and capabilities of their Autopilot and Full Self-Driving features.
Sources
ChatGPT. “What are Tesla's used technologies for Full Self Driving cars?” Accessed January 29, 2025. https://www.chat.openai.com
Wikipedia contributors. (2024, 23 juni). Tesla Dojo. Wikipedia. https://en.wikipedia.org/wiki/Tesla_Dojo