Dive Brief:
- Hyundai Motor Group is expanding its partnership with Nvidia to accelerate the deployment of next-generation, Level 2 advanced driver assist systems and more advanced Level 4 autonomous driving technology, the company announced in a press release.
- The automaker is developing a scalable AI-powered autonomous driving architecture and will integrate Nvidia’s Drive Hyperion platform in select Hyundai and Kia vehicles, including robotaxis through Hyundai's autonomous driving joint venture Motional.
- “The expanded partnership with Nvidia marks an important milestone in realizing Hyundai Motor Group’s vision for safe and reliable autonomous driving technology,” said Heung-soo Kim, executive VP and head of global strategy office of Hyundai Motor Group, in a statement.
Dive Insight:
Nvidia’s Drive Hyperion is a production-ready, end-to-end autonomous driving platform and reference architecture. It combines a standardized vehicle sensor suite, high-performance processor and software stack that supports over-the-air updates, allowing Hyundai to introduce new features and improve its performance for the entire life of the vehicle.
The platform can be used for a wide range of vehicle types, including passenger vehicles, robotaxis and delivery fleets, according to Nvidia. It supports up to 14 cameras, 12 ultrasonic sensors, nine radar units, one lidar, four interior cameras and an array of external microphones for autonomous driving.
The Hyperion platform also supports the training of AI models using continuous, large-scale data collection from vehicles operating in real-world environments. This data can be used to improve the capabilities of Hyundai’s autonomous technology over time and deliver it via an over-the-air update to all vehicles equipped with the platform. Hyundai can also use the data to develop its own proprietary AI models.
As the auto industry undergoes a shift towards SDVs, AI-powered ADAS and autonomous driving systems are widely viewed as core vehicle technologies, according to Omdia’s 2025 Software-Defined Vehicle survey released last May.
Automakers that offer their customers higher levels of automated driving are in the best near-term position to gain market share, C.J. Finn, U.S. automotive industry leader at PricewaterhouseCoopers, told WardsAuto in February.
Nvidia’s latest AI training models are part of its new open-source “Alpamayo” portfolio of simulation frameworks and physical datasets that the company unveiled in January at CES.
Alpamayo is targeted at OEMs, Tier 1 suppliers, researchers and startups to help speed up the development and deployment of autonomous driving technology. The platform's AI models also have advanced “reasoning capabilities," which can make complex driving decisions and judgments that are more like a human’s, according to Nvidia.
“The future of mobility will be built on AI and software,” said Rishi Dhall, VP of automotive at Nvidia, in a statement. “We’re combining Hyundai Motor Group’s leadership in vehicle engineering with Nvidia’s accelerated computing and AI to build safe, intelligent, Nvidia Drive-based autonomous driving systems.”
Alpamayo also supports closed-loop testing of ADAS features and autonomous driving technology in the safety of a simulated environment, allowing Hyundai to test their performance in complex urban traffic scenarios and in different weather conditions before being deployed in its vehicles. Testing in a simulated driving environment can significantly reduce development timelines and accelerate validation, according to Nvidia.
In order to prepare for the launch of future SDVs, Hyundai Motor and its sister company Kia Corp. established a new “Advanced Vehicle Platform Division” in January 2024. The unit consolidates the company’s software-defined vehicle division with its mobility engineering and technology acceleration units.
Earlier this year, Hyundai hired former Tesla engineer Minwoo Park as head of its Advanced Vehicle Platform Division. Park was one of the architects of Tesla’s Autopilot self-driving system.
