Robocar startup Tensor Auto is exploring a fundraising round of several hundred million dollars ahead of a potential initial public offering (IPO) in the US, Bloomberg reported.
The California-based company is working with Royal Bank of Canada on the process and has started approaching prospective investors, sources told the news agency.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
One option under consideration would reportedly involve raising between $300m and $400m.
Discussions remain at an early stage and the size, structure and timing of any transaction could still change.
A US IPO could take place as soon as the end of this year or in early 2027, according to the report.
Tensor Auto originated from AutoX, which launched robotaxi services in China in 2019.
The company began testing vehicles on public roads in California in 2017 and obtained a driverless permit for passenger vehicles in the state in 2020.
The company is targeting initial deliveries of its Robocars in late 2026.
According to the report, the vehicles are fitted with more than 100 sensors, including cameras, LiDAR and radars, and use artificial intelligence to process sensor data, powered by on-board chips from Nvidia.
The cars include foldable steering wheels and pedals and can be driven manually or operate autonomously.
US mobility services provider Lyft is partnering with Tensor Auto to roll out a fleet of several hundred robotaxis across Europe and North America from 2027.
Last August, Tensor announced the launch of the Tensor Robocar, which was described as the “world’s first” autonomous vehicle (AV) designed for personal ownership.
Tensor Robocar features a fully vertically integrated Level 4 autonomy stack supported by a robust sensor suite, including five LiDAR units, 37 cameras, and multiple additional sensors.
The vehicle’s AI system is modelled after human cognition, with System 1 providing instinctual responses and System 2 employing a multimodal visual language model for processing complex or unusual driving scenarios.
