Tensor's Personal Robocar and Vietnam's Emerging Role in Global Autonomy Manufacturing

Read the original Forbes article here: https://www.forbes.com/sites/bradtempleton/2025/08/13/robocar-startup-tensor-unveils-luxury-self-drive-car-for-2026/

As reported by Forbes in August 2025, Silicon Valley-based AI company Tensor introduced the world’s first personal Level 4 autonomous vehicle built for private ownership: the Tensor Robocar. While many industry players continue to focus on centralized robotaxi fleets, Tensor is exploring a different path —one where autonomy is not merely accessed via platforms, but is integrated directly into individually owned vehicles. 

Tensor's repositioning, from its earlier days as AutoX in China to its current form as a U.S.-based builder of personalized autonomous agents, reflects a broader trend toward embedded, agentic AI systems that serve individual users. The Robocar is not just a car; it is being framed as a mobile AI companion designed to enhance autonomy, privacy, and user control.  

Tensor's Robocar - Source: The Verge

About Tensor 

Founded in Silicon Valley in 2016, Tensor is an American AI company focused on building agentic technologies, products that act on behalf of their users with intelligence and autonomy. Originally known as AutoX, the company launched one of the earliest robotaxi fleets in China before divesting its operations there to concentrate on a new vision: the personal Robocar. Tensor’s flagship product embodies this shift. With its dual-mode design and advanced AI stack, the Tensor Robocar positions itself not just as an autonomous vehicle but as a mobile AI agent built for private ownership. Headquartered in San Jose, with offices in Barcelona, Dubai, and Singapore, Tensor aims to redefine how individuals engage with autonomous mobility. 

The Tensor Robocar: Redefining the technical architecture of autonomy 

The Tensor Robocar was engineered from the ground up for autonomy. With over 100 integrated sensors, including 37 cameras, 5 custom lidars, and 11 radars, the vehicle offers comprehensive situational awareness. A foldable steering wheel and retractable pedals enable seamless transitions between human and autonomous operation. 

Central to this platform is the Tensor Foundation Model, a dual-system AI framework. One part relies on imitation learning to handle real-time reflexive driving tasks, while the other uses a transformer-based visual-language model to interpret edge cases and environmental complexity. Together, they represent an effort to balance fast decision-making with broader contextual reasoning, a necessary step for consumer-grade autonomy. The Robocar was conceived from the start for private ownership, with production handled by Vietnamese automaker VinFast.

Engineering for full-stack redundancy and user autonomy 

The Robocar was designed with self-reliance at its core. Full redundancy across power, communications, and control systems, alongside sensor cleaning, diagnostics, and over-the-air updates, enables the vehicle to function with minimal human intervention. 

This level of resilience is rarely seen in consumer products. The vehicle is engineered to anticipate maintenance needs, park autonomously, and operate even in suboptimal environments, such as low-signal garages or bad weather. These capabilities hint at a shift from AVs as fleet-managed assets to self-sustaining, user-oriented machines. 

The call for visionary founders and ecosystem collaboration 

At VinVentures, we seek out founders who not only bring technical expertise but can also align with complex, multi-stakeholder environments. The Tensor case illustrates the kind of global partnership model that may become increasingly common, and necessary, as AI-native systems move into the physical world. 

As we think about the next wave of founders and frontier technologies, we are especially interested in those who can contribute to building long-term, high-trust collaborations across the global tech-industrial landscape. Apply now through: https://xzztlrf6p7q.typeform.com/VVTsApplication