VinFast is developing enhanced level 2+ and level 2++ driver assistance systems for its next-generation vehicle platform, with pilot testing of level 2++ already under way on the VF 8 and VF 9. The Vietnamese automaker is pursuing the programme through a partnership with Autobrains, an AI firm focused on automotive intelligence.

The collaboration covers two tracks: upgrading driver assistance capabilities on production vehicles, and developing a concept architecture called Robo-Car. The Robo-Car system uses seven production-grade cameras and a compact computing platform capable of processing approximately 20 trillion operations per second, foregoing LiDAR and high-definition mapping in favour of Autobrains’ Agentic AI combined with Air-to-Road localisation, which fuses live camera data with satellite imagery. The approach targets commercial scalability by reducing dependence on premium sensing hardware.

VinFast’s current electric vehicle range carries level 2 advanced driver assistance systems (ADAS) including adaptive cruise control, lane centring assist, automatic emergency braking and intelligent speed assist. The Robo-Car architecture is undergoing evaluation in controlled zones in Hanoi, with plans to extend testing to larger urban environments and international markets.

The strategy reflects a calculated emphasis on incremental validation over headline capability claims, with each development phase requiring real-world testing and safety confirmation before deployment.

Why this matters:

• Camera-first autonomy at scale is a direct commercial bet against LiDAR dependency.VinFast and Autobrains are pursuing the same sensor philosophy as Tesla and Xpeng—seven production cameras plus high-performance compute rather than expensive sensor arrays—which, if validated, dramatically lowers the per-vehicle cost floor and makes autonomous capability viable across mid-market price points that LiDAR-equipped competitors will struggle to match.
• Southeast Asia is an underserved but strategically important proving ground.Dense, chaotic urban traffic with mixed road users, inconsistent lane markings, and limited HD mapping coverage makes markets like Hanoi a more demanding test environment than the structured US cities where most AV development has occurred—and success there would be a more transferable proof point for emerging markets globally than San Francisco or Phoenix data alone.