JLR, Nvidia partner for L3 automated driving
On the way to the software-defined car and automated driving up to SAE level 3, Jaguar LandRover is enlisting the help of Nvidia. Within the framework of a strategic multi-year partnership, JLR plans to install Nvidia’s Drive platform in its cars; in addition, Nvidia will support the traditional British vehicle manufacturer with its AI expertise.
As the two companies announced at a joint virtual press briefing, JLR will install Nvidia’s Drive platform in its new cars from 2025 in all models and versions – unlike some competitors who only offer such a high-tech platform as an option, and even then only in their upmarket models. The aim of the cooperation is to develop a broad range of innovative features and functions. The focus will be on Active Safety, Automated Driving and Parking as well as ADAS. Also driver and occupant monitoring, as well as advanced visualisation of the vehicle’s environment are among the goals of the collaboration. These features are to be downloaded by users in the form of apps and installed on their vehicles even after delivery, thus enabling extensive customisation of the cars. “The age of fixed-functional vehicles is over,” said Francois Dossa, Executive Director, Strategy at Jaguar Land Rover.
Instead, the vehicles will gradually learn from their experiences via machine learning mechanisms and AI; the results will be fed back to the manufacturer, converted into corresponding control algorithms in an elaborate learning process and then regularly incorporated into the software updates. “Software-Defined cars thus will become better and better over time,” explained JLR’s Dossa. In this context, the vehicle manufacturer is developing its own operating system, JLR OS.
- The new generation of “software-defined vehicles”
- JLR developing technology-neutral vehicle architecture
The envisioned full-stack solution will be based on Nvidia Nvidia Hyperion, which features Nvidia Orin High-performance SoCs; Nvidia AV and Nvidia IX software; safety, security and networking systems; plus, surround sensors. The range of sensor technologies supported by the solution includes radar, lidar, cameras and ultrasonic sensors. According to Nvidia, According to Nvidia, automated driving on L3 using the Nvidia platform does not necessarily require that all three types of sensors are actually used. However, the greater the variety of sensors used, the better the quality of driving decisions will tend to be. Since the platform follows a modular approach and can be customized with a wide array of sensor configurations and sensor types, the company highlighted. However, neither JLR nor Nvidia commented on the question of which sensor types JLR intends to install.
Nvidia Orin is the AI brain of the car and runs the JLR OS, while Nvidia Hyperion can be regarded as the central nervous system, explained Danny Shapiro, Nvidia VP of Automotive.
Jaguar Land Rover will also leverage in-house developed data centre solutions with Nvidia DGX for training AI models and Nvidia Sim software built on Nvidia Omniverse for real-time physically accurate simulation. Jaguar Land Rover’s software-defined features and its end-to-end verification and validation architecture will enable the delivery of innovative assisted and automated driving services throughout the life of the vehicle via over-the-air software updates.
Jensen Huang, Nvidia founder and Chief Executive Officer, commented: “Next-generation cars will transform automotive into one of the largest and most advanced technology industries. Fleets of software-defined, programmable cars will offer new functionalities and services for the life of the vehicles. We are thrilled to partner with Jaguar Land Rover to reimagine the future of transportation and create the most advanced cars.”
Jaguar Land Rover aims to achieve net-zero carbon emissions across its supply chain, products and operations by 2039.