'No half-baked tech': Hyundai Motor Group vows to surpass rivals within 5 years of autonomous driving launch

Hyundai Motor Group is pouring its resources into raising the technical maturity of Atria AI, its in-house autonomous driving system. Rather than rushing to mass production to compete in a short-term launch race, the group's strategy is to accumulate sufficient data and deliver a highly refined product — a priority driven by the direct link between autonomous driving technology and user safety.

Park Min-woo, president of Hyundai Motor and Kia's AVP division and CEO of 42dot, said at Hyundai Motor Group's autonomous driving media day, held Friday at 42dot's headquarters in Pangyo, Seongnam, Gyeonggi Province, that pushing for early mass production risked yielding a mediocre Level 2++ system far short of the group's "North Star" — its ultimate goal. "Atria AI could be developed faster if we threw everything at it, but we deliberately set a later timeline," Park said.

Park also predicted that autonomous driving would soon become a basic competitive requirement for any vehicle. "Cars without autonomous driving will become difficult to sell," he said. "Just as it is hard to imagine a car without wheels, autonomous driving will become an inseparable, fundamental capability of the automobile."

Hyundai Motor Group plans to use the time remaining before Atria AI reaches mass production to sharpen functionality and safety. Kwon Jeong-hyeon, executive vice president and head of the autonomous driving development center at Hyundai Motor and Kia, said the group aims to bring driving, parking and active-safety features to a high level of product-ready completeness by 2027. "2028 will be a year dedicated to collecting vast amounts of edge-case data to build a truly safe product," Kwon said, adding that 2029 would focus on obtaining safety certifications not only in South Korea but also in North America and Europe.

The first thing Park tackled after joining Hyundai Motor Group was building a data-collection system capable of quickly identifying technical weaknesses. "The key is how fast and easily you can find high-value data — the foundation of the data flywheel," he said. "The first thing I did after taking the role was launch development of the Special Event Recorder, which logs critical situations that arise during autonomous driving."

Park particularly singled out avoiding vehicles parked or stopped on road shoulders in urban environments as a top priority for improvement. "It looks easy but is extremely difficult technically, and solving it well alone produces a significant overall quality gain," he said. "In practice, shoulder-vehicle avoidance and response to disappearing lanes have improved considerably since I joined."

Park drew a line between the group's collaboration with Nvidia and any suggestion that it was meant to replace Hyundai's proprietary technology. Hyundai Motor Group plans to begin mass-producing autonomous driving systems using Nvidia technology from 2028, while channeling the data and development experience gained in that process into further advancing Atria AI.

The group intends to convert its global production scale into a long-term data advantage in autonomous driving. Hyundai Motor and Kia sell more than 7 million vehicles a year across more than 190 countries and regions. As autonomous driving systems are fitted to a growing share of that fleet, the group will have the foundation to accumulate vast real-world road data.

"Hyundai Motor Group holds top-tier global manufacturing capabilities, which puts us in a favorable position to scale up data collection," Kwon said. "Once full-scale mass production of autonomous driving systems begins, we expect to surpass the cumulative data volumes our competitors have built up within five years."