In this blog post, we’ll examine the current state of self-driving car development and the challenges that must be overcome for full commercialization.
In 2030 South Korea, Kim Cheol-su, an employee at Company OO, gets into a self-driving car to go to work. While the car drives to the office, Mr. Cheol-su takes notes on important points and prepares for a meeting presentation. Glancing at the navigation screen, he realizes the car is taking a different route than usual, but he quickly returns to focusing on his presentation. After finishing his preparations, Mr. Cheol-su takes a short nap to relieve his fatigue and wakes up right in front of the office. The reason Mr. Cheol-su can take notes and even sleep while driving in 2030 is thanks to autonomous vehicles.
As of 2024, the global competition to develop self-driving cars is fierce. Not only automakers like Mercedes-Benz, Audi, and Tesla, but also IT companies such as Google and Apple, as well as graphics processing specialists like NVIDIA, are actively engaged in developing self-driving technology. Each company is utilizing information processing technology to develop self-driving systems that enable cars to independently assess their surroundings, calculate driving routes, and arrive at their destinations safely and quickly.
The core of autonomous driving technology lies in the ability to recognize the vehicle’s surroundings. To this end, environmental recognition technologies utilizing sensors and graphics processing units (GPUs) have been developed and are being integrated into autonomous vehicles. Sensor-based environmental recognition technology detects obstacles in the vicinity through various sensors mounted on the vehicle, enabling the vehicle to accurately assess its surroundings. The key to sensor-based technology is reducing blind spots by utilizing a variety of sensor equipment. On the other hand, technology utilizing graphics processing units involves analyzing video footage captured by cameras mounted on the vehicle through an image processor to understand the vehicle’s surroundings. The cameras act like human eyes, and the processor acts like a brain, enabling video analysis through big data and machine learning.
Although autonomous vehicles are being developed using various technologies, fully autonomous driving has not yet been achieved. Currently, commercialized autonomous driving technology is limited to maintaining lanes on the road and driving in accordance with traffic flow. However, if technological development continues, fully autonomous driving will become possible in the future, allowing us to use driverless cars just like in the movies *Minority Report* or *I, Robot*. This will enable various activities inside the car, such as working, reading books, or watching movies, transforming the car from a space for tiring driving into an office or a place of rest.
Nevertheless, there are many issues that must be resolved before the era of fully autonomous driving arrives. With the emergence of this new technology, clear social discussion and the establishment of relevant laws are necessary. If a traffic accident occurs involving an autonomous vehicle, who would be considered at fault? Since machines are unlikely to malfunction unless they break down, does that automatically mean the human is at fault? Or should the party at fault be determined by considering the probability of machine malfunction? If self-driving cars are allowed on the roads without addressing these issues, society will be engulfed in heated debate.
Verifying the safety of self-driving cars is also essential, as it is directly linked to the safety of the passengers. In fact, there was a recent incident where a Tesla vehicle, driving with its autonomous driving feature activated, collided with a trailer, resulting in the driver’s death. There is also a case involving Google’s self-driving car, currently under development, which was unable to avoid a collision with a vehicle that ran a red light at an intersection. If we blindly trust imperfect autonomous driving technology, we may fail to prevent accidents that humans could have easily avoided, leading to massive social losses.
Governments and companies in various countries are continuing discussions to address this issue. Among them, the actions taken by the United States—one of the world’s largest automotive markets—are particularly noteworthy. In June 2011, Nevada became the first state in the world to allow autonomous vehicles on the road, and the first test licenses for autonomous driving technology were issued. Recently, under the leadership of the federal government, guidelines for autonomous vehicles applicable across the entire United States were announced.
South Korea also plans to establish special zones for autonomous vehicles, led by the Ministry of Land, Infrastructure and Transport and the Ministry of Trade, Industry and Energy. However, it is regrettable that, as the world’s fifth-largest automobile producer, the country is moving so slowly to embrace this crucial future industry. The U.S. federal government’s guidelines will be of great help to South Korea in preparing swiftly for the era of autonomous vehicles.
While the commercialization and widespread adoption of autonomous vehicles may vary depending on various factors, current projections suggest that fully autonomous vehicles (Level 5) will not become mainstream until at least 2035. This is because more time is needed to resolve technical, regulatory, and safety issues. Autonomous vehicles are an attractive technology, and if utilized appropriately, they could significantly improve the quality of life for busy modern people. However, if we adopt the technology without first conducting sufficient social discussion, enacting appropriate regulations, and verifying safety, social conflicts are likely to escalate. There are still ten years remaining until 2035, and technological advancements are occurring rapidly. I hope that during this time, our society will adequately prepare for autonomous vehicles to enhance the convenience of all its members.
