ADVERTISEMENT – CONTINUE READING BELOW
Nvidia now has the intelligence needed to scale up self-driving, software-defined vehicles with over-the-air update capability. This is the future of mobility. All automakers are working on autonomous driving, which manifests itself in various levels from simple cruise control or automatic emergency braking to hands-free highway driving and robotaxis.
Nvidia can give an automaker an AI platform so they don’t have to build their own from scratch, allowing for quicker time to market. Mercedes-Benz did just that: The 2026 CLA introduces the company’s latest operating system, MB.OS, which enables communication between the car’s sensors and brain, running everything from infotainment to hands-free driving. Mercedes worked with Nvidia to bring the AI piece of the puzzle to the car.
ADVERTISEMENT – CONTINUE READING BELOW
AI also forms the backbone of Drive Assist Pro, the Mercedes equivalent of Tesla’s Full Self-Driving, but with a super-redundant software stack for safety. Mercedes built it with Nvidia using a combination of real-world and virtual data, creating scenarios and simulations to turn a single incident into hundreds of thousands of miles of test driving. By working with Nvidia, Mercedes got the system up and running sooner than if the automaker did all the work in-house.
The problem the industry has faced to date is that AI, while a quick study, still does not know how to handle every one-off scenario a vehicle may encounter. The challenge is safely navigating the rare and complex situations that fall outside the training experience.
This is where Nvidia looks to revolutionize the game. In January it launched Alpamayo, a family of AI models and simulation tools and datasets that can train robots and vehicles—in other words, physical machines—to think like a human to sort out complex driving scenarios.
Automakers can put Alpamayo in their autonomous vehicle stacks as a teacher model and then fine tune. This should accelerate safe autonomous driving, pushing it into the mainstream. The package aims to help automakers develop vehicles that perceive, reason, and act like humans to solve problems, like how to navigate a broken traffic light at a busy intersection without previous experience. Self-driving cars will possess the ability to analyze and act in the real world by breaking down a problem into steps, reasoning through every possibility, and selecting the best course of action.