[Edaily Reporter Lee So-Hyun ] SOCAR(403550)and Apex Mobility are launching a full-scale business by positioning “autonomous driving data” itself—rather than vehicles or sensors—as their core product. The plan is to process the real-world driving and accident data accumulated by SOCAR over the past 15 years into a format suitable for artificial intelligence (AI) training and supply it to domestic and international autonomous driving companies, automakers, and research institutions.

SOCAR and Apex Mobility announced on the 24th that they will participate in the “2026 Autonomous Mobility Industry Expo (AME 2026),” taking place from the 25th to the 27th at COEX in Samseong-dong, Seoul. The two companies plan to use their exhibition space as a platform to showcase their competitive edge in autonomous driving data and as a venue for external consultations aimed at expanding their data business.

The theme of this exhibition is “Korea’s Largest Autonomous Driving Data Platform.” This stems from the assessment that, as autonomous driving technology shifts from a rule-based approach relying on high-resolution precision maps to an end-to-end (E2E) approach that processes multiple data sources—such as camera footage and LiDAR—through a single AI model, the focus of technological competition is shifting toward data.

Collecting Real-World Driving Data from 25,000 Vehicles Nationwide

All approximately 25,000 car-sharing vehicles operated by SOCAR nationwide are connected cars. More than 100 types of vehicle data—including dashcam footage from front and rear dual-channel cameras and GPS and steering angle data—are collected in real time via the company’s proprietary telematics terminals.

According to SOCAR, the total daily mileage of its fleet is approximately 1.1 million km. This is 10 times the total length of South Korea’s roads (approximately 110,000 km), enabling the company to capture data on actual public road conditions 24 hours a day, 365 days a year, across all four seasons.

In particular, SOCAR emphasized its strength in “edge case” data, which is essential for improving the reliability of autonomous driving AI. By continuously collecting over 40,000 accident videos and e-motion data from the moment of the accident each year, the company has accumulated approximately 220,000 related data points to date.

Apex Mobility processes the raw data—through time-syncing and automatic labeling of driving situations using a Vision Language Model (VLM)—into an international standard format that can be immediately utilized for autonomous driving simulations and other applications.

At this exhibition, Apex Mobility will also unveil for the first time a prototype vehicle equipped with various sensors to enable the company to independently collect high-quality autonomous driving data. The prototype Ioniq 5 is equipped with a LiDAR sensor on its roof and seven cameras mounted throughout the vehicle to capture high-resolution, multi-angle video from all directions.

On-Demand Data Collection Also Available

SOCAR and Apex Mobility also offer a custom data collection feature that collects and supplies data tailored to a client’s specific needs—including location, timeframe, vehicle type, and scenario.

SOCAR explained that since it directly owns the vehicles and manages their maintenance, deployment, and retrieval, it can accumulate real-time data without the need for separate data collection vehicles or dedicated drivers. The installation of onboard units and software deployment can also be carried out internally.

Apex Mobility plans to process not only the data collected from SOCAR vehicles but also the high-quality data it will collect independently through the future operation of vehicles equipped with full sensor suites, all in accordance with international standards. Through this, the company aims to provide training data tailored to the specific conditions requested by clients, such as specific regions, driving situations, and vehicle types.

The two companies emphasized their competitive edge in the data business, stating, “There are very few operators in the global market that possess all three elements necessary for training end-to-end (E2E) autonomous driving models: a large fleet of vehicles, data collection capabilities, and a centralized data pipeline.”

Park Jae-wook, CEO of SOCAR and Apex Mobility, said, “In the era of end-to-end autonomous driving, where AI makes decisions on its own, the volume and quality of data from real roads will be the deciding factor.” He added, “We have a real-time, large-scale data pipeline that no other domestic mobility platform can match, and we plan to begin full-scale discussions with domestic and international companies regarding the supply of autonomous driving data.”

Last May, SOCAR established Apex Mobility, a joint venture for autonomous driving with a total investment of 150 billion won, and launched related business operations. Apex Mobility plans to begin by building an autonomous driving data pipeline and establishing supply partnerships, and will gradually expand its operations to include autonomous driving-based robo-taxi and ride-hailing services in the future.