Socar will jointly develop AI tire-reading technology with vehicle inspection AI solutions firm Autopedia. Socar said on Tuesday it signed a memorandum of understanding with Autopedia for the joint development of AI tire-reading technology.

The two companies have managed tire wear on Socar vehicles using Autopedia’s AI solution, Doctor Tread, since July last year. From January to April, the rate of tire-related breakdowns fell 22 percent from a year earlier. Over the same period, the number of customer inconvenience cases declined 20 percent. The agreement was designed to expand the scope and precision of readings based on those results.

The next-generation model aims to go beyond wear readings to precisely identify uneven wear, punctures caused by foreign objects such as nails and cuts, and chunking, or rubber tearing. The plan is to build a system that detects early, subtle warning signs with AI that are difficult to capture with visual inspection or single-indicator readings, and to carry out preventative maintenance before breakdowns occur.

Both companies will participate in the entire development process, including model design, data labeling and performance verification. Socar will provide real tire image data accumulated through its car-sharing operations and reading standards from maintenance sites. It plans to enhance the precision reading model by adding Autopedia’s AI capabilities.

Moon Jeong-woong (문정웅), head of Socar’s car-sharing customer experience division, said tire management is directly linked to customer safety, making precise, constant checks and preventative action critical. He said the company will work with Autopedia to find failure factors that are difficult to confirm with the naked eye at an early stage and strengthen safety management so customers can use Socar with peace of mind at any time.

Kim Sang-yeon (김상연), vice president of Autopedia, said Doctor Tread is a solution Autopedia has independently developed over a long period and is based on more than 100,000 tire data points, making it a top-level domestic technology. He said the companies will enhance the reading model to develop technology that strengthens safety and will work to make it a benchmark for the domestic autonomous maintenance market.