Tensor is positioning its Robocar as a privately owned SAE Level 4 vehicle, engineered around autonomy rather than retrofitting sensors onto an existing platform. The design is sensor-first: 5 LiDAR units, 37 cameras, 11 radars, plus microphones and underbody detection to see close to the curb and avoid low obstacles, with a cleaning system (large fluid tank, air/liquid jets, wipers) to keep optics usable in real-world grime. https://www.tensor.auto/
A big theme is fail-operational redundancy: braking, steering, power and compute are treated as duplicated subsystems, with partners mentioned like Bosch, ZF and Autoliv for safety-critical hardware. Tensor’s approach relies on multi-modal sensor fusion—using the strengths of vision, radar and LiDAR together—so the stack can handle edge cases like occlusion, glare, and near-field perception without betting everything on a single modality, which is where many autonomy programs see risk.
The interview was filmed at CES Las Vegas 2026, where Tensor also talked about opening parts of its AI work to outside developers. Beyond the car itself, they point to open tooling for “physical AI” workflows (vision-language-action training and deployment), and say the core models are being released in an open form, inviting collaboration while keeping the vehicle’s runtime data local to the car, via OpenTau.
Inside, the cabin is treated like a productivity and media space: multiple displays, individual in-cabin cameras for calls, and privacy shutters for sensor coverage you want to disable. The signature mechanical element is a fold-away steering wheel and pedals that pop out on demand, making the handoff between Level 4 autonomy and manual control explicit, and supporting a spectrum from Level 3/2 ADAS down to Level 0 for fully human driving mode.
On go-to-market, Tensor frames a hybrid of personal ownership and fleet economics: owners can optionally connect the vehicle to ride-hailing when idle, while fleet partners like Lyft and the Copenhagen car-sharing operator GreenMobility have been announced as early channels. Manufacturing is planned via VinFast in Vietnam, with production targeted for the second half of 2026 and deployments likely constrained to geofenced ODD areas before broader roll-out in 2026.
Tensor Robocar Level-4 autonomy: 100+ sensors, Nvidia Thor compute, dual-mode cabin
Tensor autonomous car: LiDAR/radar/camera fusion, retractable wheel, privacy-first on-device AI
I’m publishing about 100+ videos from CES 2026, I upload about 4 videos per day at 5AM/11AM/5PM/11PM CET/EST. Check out all my CES 2026 videos in my playlist here: https://www.youtube.com/playlist?list=PL7xXqJFxvYvjaMwKMgLb6ja_yZuano19e
This video was filmed using the DJI Pocket 3 ($669 at https://amzn.to/4aMpKIC using the dual wireless DJI Mic 2 microphones with the DJI lapel microphone https://amzn.to/3XIj3l8 ), watch all my DJI Pocket 3 videos here https://www.youtube.com/playlist?list=PL7xXqJFxvYvhDlWIAxm_pR9dp7ArSkhKK
Click the “Super Thanks” button below the video to send a highlighted comment under the video! Brands I film are welcome to support my work in this way 😁
Check out my video with Daylight Computer about their revolutionary Sunlight Readable Transflective LCD Display for Healthy Learning: https://www.youtube.com/watch?v=U98RuxkFDYY
