Kakao Mobility’s Seoul Autonomous Car Travels 13,000 km Across Gangnam in Six Months

Over 2,000 Completed Rides Recorded
Average User Rating: 4.97 out of 5

'Seoul Autonomous Car,' a self-driving taxi service operating in the Gangnam area of Seoul during nighttime hours, has achieved a total driving distance of 13,000 km after just over six months since its launch.

On the morning of September 9 in Seocho-gu, Seoul, Minseon Kim, Head of the Autonomous Driving Business Team at Kakao Mobility, told reporters, "The cumulative driving distance of the Seoul Autonomous Car has exceeded 13,000 km, and the number of completed rides has surpassed 2,000."

On the morning of the 9th, the autonomous taxi Seoul Autonomous Car, operated by Kakao Mobility, was exhibited at the Korea Automobile Mobility Industry Association in Seocho-gu, Seoul. Photo by Myunghwan Lee

The average distance per ride for users was recorded at 2.9 km, with an average travel time of 9.7 minutes. User satisfaction was also reported to be high, with the service achieving a score of 4.97 out of 5 in user ratings, based on feedback from 235 participants.

Kakao Mobility was selected as the operator for Seoul's autonomous vehicle passenger transport service and began late-night operations (from 10:00 p.m. to 5:00 a.m. the following day) in the Gangnam area in March. Since last month, the number of vehicles in service has increased from two to six.

No accidents have occurred in which the Seoul Autonomous Car was at fault, further demonstrating its safety. Kim emphasized, "There was a case where a stationary vehicle was lightly rear-ended by another car during the operating period, but there were no personal injuries." He added, "A dedicated team is always on standby to respond immediately to any incidents."

Over the past six months, while operating the Seoul Autonomous Car, Kakao Mobility has collected driving trajectory data for the entire Gangnam district, and since June, has also started accumulating daytime data when traffic volume is highest. As the Gangnam area experiences heavy traffic even at night and frequent unexpected situations, such as pedestrians suddenly entering the road or sudden movements within alleyways, the company is continuously collecting data to respond to these scenarios as well. All collected driving data undergoes processes of collection, refinement, and labeling, along with retraining of models, and is utilized to further advance autonomous driving technology.

Currently, Kakao Mobility is applying both an artificial intelligence (AI) planner and a rule-based planner within its autonomous driving system. By the end of this year, the company plans to increase the weight of its E2E (End-to-End) autonomous driving model, which processes everything from perception to decision-making in a single model. At present, the company is testing the performance of E2E model-based driving plans on actual roads.

Inho Lim, leader of the AI Driving section of the Autonomous Driving Development Team, said, "Through the Seoul Autonomous Car service, we have extended the AI planner into an E2E model," adding, "We have supplemented it with a rule-based safety evaluation and are verifying its performance in real-world road conditions."

Lim also emphasized, "To truly commercialize driverless robot taxis, it is essential to secure the mobility platform, autonomous driving technology, and the related ecosystem. Kakao Mobility is steadily assembling all three components."

To use the Seoul Autonomous Car, users can select the taxi icon in the Kakao T app during operating hours, enter their departure and destination points within the Gangnam-gu service area, and then choose 'Seoul Autonomous Car' from the call menu. If a Seoul Autonomous Car is unavailable, users are automatically connected to an alternative means of transportation, such as a regular taxi.

This content was produced with the assistance of AI translation services.