This presentation explores recent advancements in AI hardware and techniques that are transforming automotive technology. At Magna, we are focusing on two key areas. First, using Large Language Models (LLMs) to enhance interior cabin interactions, allowing natural voice commands like “I’m cold” to control climate settings seamlessly. Second, improving autonomous driving by applying innovative approaches, including diffusion models for trajectory planning and working toward “End-to-end Driving” systems. While perception systems have become increasingly AI-based, ADAS features like adaptive cruise control are still often implemented using traditional programming techniques. With new hardware and methods, we are working to modernize this layer for better performance and safety.
This talk will share initial results and lessons learned, demonstrating how these innovations can create more intuitive, reliable, and capable vehicles in the near future.
