Autobrains, an Israel-based AI company, has begun deploying what it calls the automotive industry’s first Agentic AI architecture for Advanced Driver Assistance Systems (ADAS) and automated driving. The technology is now in production programs with global automotive partners on mass-market vehicles using standard sensor configurations.
Highlights
- Agentic AI architecture uses specialized, scenario-focused driving agents instead of a single monolithic end-to-end model
- Selective activation engages only the agents relevant to a given driving situation, reducing compute load on existing vehicle hardware
- Mass-market deployment targets standard automotive-grade sensors and compute platforms, avoiding costly hardware upgrades
- OEM partnerships span L2+, L2++, and higher-level automated driving programs with leading automakers and Tier 1 suppliers
How Agentic AI Differs From Monolithic Approaches
As ADAS and automated driving systems have advanced, many have grown more expensive to scale. Monolithic end-to-end architectures typically rely on increasingly large models and expanding data requirements. Autobrains takes a different approach by organizing driving intelligence into multiple specialized agents.
Each Driving Agent handles specific scenarios. Only agents relevant to the current driving situation activate at any given moment. This selective activation reduces compute demand and allows capability expansion within existing vehicle platforms.
Implications for OEMs
For automakers, the architecture removes the need to redesign vehicle platforms to unlock new safety or autonomy features. Instead, capabilities can be delivered through software updates on hardware already installed in mass-market vehicles.
“Autobrains’ Agentic AI approach enables OEMs to evolve driving capability within existing vehicle platforms,” said Igal Raichelgauz, CEO of Autobrains. “It gives OEMs architectural control over a system designed for continuous capability expansion rather than a fixed, monolithic software stack.”
Raichelgauz added that autonomy will not scale through additional hardware alone. “It will scale by organizing intelligence differently,” he said.
Backing and Production Programs
Autobrains has raised more than $140 million from investors including BMW, Toyota Ventures, VinFast, Continental, Knorr-Bremse, Magna, and Temasek. The company holds over 300 patents in AI and autonomous driving. Its production programs include design wins with leading OEMs and Tier 1 suppliers across multiple levels of driving automation.
Sign up for our weekly emails to stay up to date on the latest news!
