▲Lee Jae-kwan, head of the Autonomous Driving Technology Research Center at the Korea Automobile Research Institute, is giving a presentation.


Active development of autonomous driving-related technologies such as simulation, mapping, and V2X
Foxconn Develops Autonomous Driving with Open Platform… 2,748 Companies Participate

Experts continued to make presentations about autonomous driving, which has not yet arrived due to safety issues such as edge cases, and that active research and development is still being conducted as it is called the future of the automobile industry.

Jae-Kwan Lee, head of the Autonomous Driving Technology Research Center at the Korea Automobile Research Institute, gave a keynote speech on the topic of “The Future and Reality of Autonomous Driving” at the “MATLAB EXPO 2024 Korea” held by MathWorks at the Grand InterContinental Seoul Parnas on the 11th, and shared trends and major issues in the autonomous driving industry.

When Honda announced its Level 3 autonomous vehicle, the Legend, in 2021, it emphasized that it would never release it without ensuring safety.

As much as they emphasize the safety of autonomous vehicles, they restrict various functions, such as driving at speeds of less than 50 km/h, and they are not sold to the general public. Instead, only 100 units were delivered to the government and then discontinued.

The director said, “If an error occurs in the perception, judgment, and control that are the core of the autonomous driving system in a high-speed vehicle, it can become a moving weapon.”“This is also the reason why European automotive companies are placing considerable emphasis on safety and security in ISO TC 2,” he said.

'Edge cases', which refer to unexpected scenarios, are a major obstacle to autonomous driving.

Last year, a pedestrian who had collided with a third vehicle entered the Cruise vehicle's lane, but the Cruise vehicle, unaware of this, pushed the pedestrian out of the way, causing a secondary accident.

The director said, “Argo AI, which attracted a lot of attention with Ford and Volkswagen investing about 6 trillion won, also had a hard time responding to both standard and non-standard objects as well as scenarios that could occur on the road,” and said that the wall to autonomous driving is by no means low.

He then predicted that although the autonomous driving industry is facing many challenges, autonomous driving cannot be excluded at this time of change in the automobile industry.

About 10 years ago, it was predicted that annual vehicle sales would reach 160 million units, but it is only about 80 million units.

The old automobile industry emphasized volume, but with sales that have declined since the coronavirus pandemic and are not expected to increase again, companies have started looking for other sources of income, and autonomous driving is one of them.

It is also worth noting the research and development of companies and the emergence of new players to take the lead in the new food market.

A representative example is the method of shortening the safety test period, which originally took 12 years, through simulation as highlighted above.

Not only simple autonomous driving systems, but also sensor error and physical modeling of surround sensors are done through simulation.
<bResearch and development on mapping, such as cloud digital maps and dynamic maps, is also active.

It is known that Naver is investing about 3 trillion won and Cruise is investing about 2 trillion won in this technology.

This is a technology that helps make autonomous driving safer by adding layers such as △Road Traffic Act weather forecasts on top of the basic map layer, △layers containing information such as accident occurrence and current weather information, and △layers containing information such as surrounding vehicles and pedestrians using V2X.

The V2X (Vehicle to Everything) mentioned above refers to technology that communicates with various objects on the road for autonomous driving.

V2X is attracting attention because it can improve safety by wirelessly transmitting information about pedestrians, motorcycles, etc. in blind spots that sensors cannot detect.

Director Lee Jae-kwan selected Foxconn as a notable company in the autonomous driving industry.

Foxconn, known as a contract manufacturer for Apple products, is developing an electric-based autonomous vehicle with the goal of launching it in 2025-2026.

Apple has given up on the Apple Car, but it is expected that if Foxconn makes a self-driving car, it will continue its business by supplying content for the vehicle.

The industry sees Apple's decision to abandon the Apple Car as a delay in re-entering the market as technology matures, rather than as an abandonment of the company's 10-year research.

The reason Foxconn is worth paying attention to is that it has declared that it will develop into an open platform for global players by turning Taiwan's weakness of not having an automobile value chain into an advantage.

It is noteworthy that 2,748 members are participating in Foxconn's development plan, including top companies in the autonomous driving and electric vehicle industries such as Nvidia, LG Energy Solution, Infineon, and Valeo. There is a need to do it.

Kim Pil-soo, a professor in the Department of Future Automobile Engineering at Daerim University, also predicted the future of Foxconn in an article he contributed in May, saying, “The future electric vehicle that Foxconn is pursuing is an ‘electric vehicle foundry’ that creates an electric vehicle with a thousand faces by putting a cover on the platform and changing the algorithm,” and “A new electric vehicle is created by applying a company’s specialized algorithm, and it is a future mobility that includes autonomous driving algorithms including generative artificial intelligence.”