Mars Auto, a developer of autonomous driving solutions for heavy-duty freight trucks, announced that it will launch a Tesla-style autonomous driving software system later this year, outlining plans to lead autonomous logistics routes connecting South Korea and the United States. The company presented its vision at a media event in Seoul on July 1, stating that it aims to fully automate freight transportation from South Korea to the U.S. using end-to-end (E2E) artificial intelligence (AI)-based trailer autonomous driving technology. Park Il-soo, Chief Executive Officer (CEO) of Mars Auto, said the company aims to commercialize driverless trucks by 2028. He noted that Mars Auto will focus on developing an E2E autonomous driving system optimized for trucks, relying primarily on camera-based perception rather than LiDAR or other sensor-heavy approaches. CEO Park, who founded Mars Auto in 2017, said he targeted what he called one of the "most difficult problems solvable by AI." While exploring business opportunities at the time, he was influenced by Waymo LLC’s efforts in autonomous trucking and saw strong potential in freight transport alongside robotaxi applications. He added that autonomous trucking offers a faster path to commercialization because approximately 98% of long-haul truck driving occurs on highways, where technological implementation is comparatively easier. In particular, CEO Park emphasized that Mars Auto focuses on a camera-based approach similar to Tesla Inc.’s, enabling AI systems to interpret visual data and make driving decisions autonomously. He noted that this approach offers both technological performance and significant cost advantages. The company stated that while competitors require roughly 35 million won per vehicle in system deployment costs, Mars Auto’s solution can be implemented at around 1 million won per vehicle. To build its dataset, Mars Auto initially collected driving data using the simulation game “Euro Truck,” which replicates European trucking environments. The company later