Nissan has announced that it will use self-driving software developed by Wayve to enhance its advanced driver assistance system starting in 2027. This is a landmark deal for the high-profile British startup, which has raised more than $1.3 billion from Nvidia, Microsoft, SoftBank Group and Uber.
Nissan will integrate Wayve’s software as well as sensors, including lidar from an unknown supplier, into its ADAS system, known as ProPilot. The Japanese automaker said the system will set a “new standard for autonomous driving with advanced collision avoidance capabilities.”
Nissan said that the system, which is being developed for mass production by 2027, will be a Level 2 system that works under the supervision of the driver. Nissan did not disclose in which models this system will be available. Level 2, as defined by the Society of Automotive Engineers, means that the system can automate several driving tasks, such as steering and braking, but still requires constant monitoring by the driver.
Wayve, which was launched in 2017, has attracted attention and investors for its automated driving system that is self-learning rather than rule-based, similar to Tesla’s approach. Wayve has developed its end-to-end data-driven learning approach for “on” assisted driving and a fully automated “eyes-free” driving system.
Unlike Tesla, Wayve plans to sell its “Embodied AI” to automakers and other technology companies.
Wayve’s self-learning approach is considered to be particularly attractive to automakers because it is not dependent on specific sensors or HD maps. This means that the Wayve system can work with existing sensors such as cameras and radars. The automated driving software collects data from these sensors, which directly affects the system’s driving decisions.
And while the startup’s development fleet uses a system on Nvidia’s Orin chip, co-founder and CEO Alex Kendall said the software can run on any GPU that is already installed in its OEM partners’ vehicles.
All of this means that the software will be cheap to use and will be used in advanced driver assistance systems, robotaxis, and even robotics, Kendall said last month at the Nvidia GTC conference.
Nissan spokesman Shiro Nagai said in an email that Wayve’s Embodied AI base model – large-scale end-to-end AI trained on a huge amount of real-world driving data – allows the software to adapt across environments and vehicle platforms.
“By leveraging Nissan’s technology and Wayve’s AI, we are confident that it will enable future Nissan vehicles to accurately replicate the judgment and actions of a careful and competent driver in complex driving scenarios,” Nagai said.
