
Kakao Mobility will unveil for the first time this year a self-driving service based on "end-to-end (E2E)," a next-generation autonomous driving technology that integrates and processes camera and sensor information.
"We trained the E2E model based on data obtained from actual driving, and we have created a verification process and are conducting data verification," Kim Jin-kyu, vice president of Kakao Mobility's Physical AI Division, said on the 28th at a debate titled "In the Age of Physical AI, How Does Autonomous Driving Become Part of Daily Life," held at the National Assembly Members' Office Building in Yeouido, Seoul. "If possible, we will be able to deploy it in the Gangnam service by the end of the year."
E2E is a technology in which an artificial intelligence (AI) model learns from road images, maps, and sensor data acquired during driving to make its own judgments and determine how to control the vehicle. Tesla has already switched its full self-driving (FSD) to the E2E method, advancing its technology and moving toward commercialization. As a latecomer, Kakao Mobility is working to internalize the core E2E autonomous driving technology, in which a single integrated AI model performs everything from perception to control.
"Tesla recently demonstrated that going with the E2E method works, and Chinese companies are also rapidly following the same approach," Kim said. "Kakao Mobility is also swiftly transitioning from a rule-based method toward the E2E direction."
Since March, Kakao Mobility has been operating seven autonomous vehicles under the "Seoul Autonomous Vehicle Service" in the Gangnam district. It has also signed a memorandum of understanding with LG Innotek and Kia, aiming to develop a dedicated autonomous vehicle.
"Autonomous driving is a representative field of Physical AI in which AI provides services to citizens on actual roads and in cities," said Ryu Keung-sun, CEO of Kakao Mobility. "As important as the technology are the capabilities across overall service operations, including the platform, control, data utilization, safety management, and user experience."
Ryu introduced cases of technology development and accumulated demonstration experience in Seoul, Sejong, and Pangyo since Kakao Mobility launched its autonomous driving task force (TF) in 2018. "This technology development experience and platform operation capability will be an important foundation for autonomous driving to move beyond the demonstration stage and establish itself as a transportation service in daily life," he said.






