VinFast and Autobrains Forge Strategic Partnership to Democratize Autonomous Driving

HANOI, VIETNAM & TEL AVIV, ISRAEL – In a move set to reshape the economic landscape of self-driving technology, VinFast and Autobrains have announced a strategic partnership today, January 27, 2026. This collaboration aims to develop an affordable Level 2++ (L2++) autonomous driving solution and a revolutionary "Robo-Car" system. By leveraging Autobrains' proprietary AI architecture, the partnership seeks to eliminate the industry's reliance on expensive sensors, marking a significant pivot toward accessible, mass-market autonomous mobility.

The alliance brings together VinFast, Vietnam’s leading electric vehicle (EV) manufacturer, and Autobrains, an Israeli AI innovator known for its paradigm-shifting "Liquid AI" technology. Together, they are targeting a critical bottleneck in the autonomous vehicle (AV) industry: cost. While traditional AV systems rely on prohibitive hardware suites, this new initiative promises to deliver high-level autonomy using a camera-first approach, effectively lowering the barrier to entry for advanced driver-assistance systems (ADAS) globally.

The "Robo-Car" Architecture: A Camera-First Revolution

At the heart of this partnership is the development of a new "Robo-Car" architecture. Contrary to the prevailing industry trend that bundles LiDAR, radar arrays, and high-definition maps into complex and costly sensor suites, the VinFast-Autobrains solution utilizes a streamlined configuration.

The system relies on seven standard cameras paired with a compact, high-performance computing chip. Despite its lean hardware footprint, the system is capable of processing approximately 20 trillion operations per second (TOPS). This efficiency allows the vehicle to achieve advanced spatial awareness and decision-making capabilities without the hardware bloat that typically drives up the price of autonomous EVs.

This architectural shift aligns with a growing school of thought in AI development—championed by industry disruptors—that visual data, when processed by sufficiently advanced AI, is superior to active sensor fusion for general driving tasks. By removing LiDAR, VinFast projects a dramatic reduction in production costs, a saving that can be passed directly to the consumer to make L2++ features standard rather than luxury add-ons.

Unleashing "Liquid AI": A New Approach to Machine Learning

The technological backbone of this system is Autobrains' Liquid AI. Unlike traditional deep learning models, which require massive datasets of labeled images and struggle with "edge cases" (rare, unpredictable driving scenarios), Liquid AI operates on a signature-based self-learning framework.

Key differentiators of Liquid AI include:

• Unsupervised Learning: The system maps raw, real-world data to compressed signatures, allowing it to "learn" driving concepts without extensive human labeling.
• Modular "Skills": The driving problem is decomposed into smaller, context-specific AI models called "Skills." These skills are activated dynamically based on the scenario—whether it's merging onto a highway or navigating a crowded urban intersection.
• Edge Case Adaptability: Because the system is not rigidly trained on a fixed dataset, it adapts more fluidly to unexpected road conditions, a critical safety factor for L2++ and L3 autonomy.

This "Agentic AI" architecture enables the vehicle to perceive and react with near-human precision. It addresses the "perception-decision gap" often found in black-box neural networks, providing a more transparent and explainable decision-making process.

Strategic Roadmap: From Pilot to Mass Market

The collaboration is already yielding tangible results. VinFast has commenced pilot testing of the upgraded L2++ system on its flagship VF 8 and VF 9 electric SUVs. These tests, currently conducted in controlled zones in Hanoi, focus on validating the system's ability to handle complex urban environments using the new camera-only protocol.

The roadmap extends beyond immediate upgrades. The ultimate goal is to integrate the Robo-Car system across VinFast's entire product portfolio. By standardizing this technology, VinFast aims to differentiate itself in the crowded EV market not just through battery range or design, but through superior, accessible intelligence.

Projected Rollout Phases:

1. Phase 1 (Current): Pilot testing of L2++ on VF 8 and VF 9 in Vietnam.
2. Phase 2: Expansion of testing to larger international cities to validate Liquid AI against diverse traffic laws and driving behaviors.
3. Phase 3: Commercial integration of the Robo-Car architecture into future VinFast models, aiming for a global release of affordable self-driving capabilities.

Comparative Analysis: Traditional vs. Liquid AI Architectures

To understand the magnitude of this shift, it is essential to compare the Autobrains approach with the current industry standard for L2++/L3 autonomy.

Table 1: Technical Comparison of Autonomous Driving Approaches

Industry Implications and Future Outlook

The partnership between VinFast and Autobrains signals a potential "democratization moment" for the autonomous vehicle sector. For years, the narrative has been that safer cars require more expensive eyes. By challenging this axiom, VinFast is positioning itself as a pioneer of affordable robotics.

For the broader AI industry, this validates the shift away from brute-force data crunching toward more elegant, biomimetic architectures. If Autobrains' Liquid AI proves successful in the diverse and chaotic traffic conditions of Vietnam, it will serve as a robust proof-of-concept for camera-based autonomy worldwide.

Creati.ai will continue to monitor the progress of the VF 8 and VF 9 pilot programs. As the industry watches closely, the success of this "frugal autonomy" could force a reassessment of hardware strategies among legacy automakers and tech giants alike.