Autonomous Vehicle, Weekly News #17 / 2024
※ This article summarizes interesting news related to autonomous driving mobility not only in Korea(mainly) but also in other countries every week. I put some my comments in italics, the news related to South Korea may contain links and documents in Korean.
<< South Korea >>
There are limited areas where autonomous vehicles can be found in Korea. Still, related tests and pilot operations are already underway across the country. And the type of transportation that has recently been attracting attention in many regions is public transportation. In particular, we are making efforts to provide people with the opportunity to experience autonomous technology through autonomous buses. Of course, they are currently also providing services through robotaxi. Still, they are expecting practical results by deploying autonomous buses on existing bus routes. In addition, plans are being released to apply autonomous driving technology beyond large buses to village buses operating in small areas. Perhaps it will be introduced next year. Autonomous village buses are expected to operate on narrow roads rather than existing bus routes. Because it requires safer and more precise driving capabilities. Perhaps it would be operated by shuttle bus.
In any case, compared to the past, Korean companies developing autonomous driving technology have also made less investment. However, the government is preparing for the commercial emergence of fully autonomous driving in 2027 through consistent selection of autonomous vehicle pilot operation zone (AV POZ). Through this, new startups are also emerging. I believe that it is important for Korean companies to secure software-centered technology that can overcome the limitations of Korea’s narrow market.
[Company]
▶ Hyundai Mobis: Hyundai Mobis tests Level 4 autonomy
Hyundai Mobis Co. will start test driving a Level 4 autonomous driving vehicle in Incheon’s Songdo and Yeongjongdo from April to next year.
The company signed a business agreement with the Incheon Free Economic Zone Authority and Incheon Technopark for commercializing autonomous driving technologies by testing long-distance urban routes.
The test vehicle, equipped with advanced computing technology for Level 4 autonomy, will cover a 60-kilometer route centered around the Incheon Bridge. The company explained that this distance is long for urban areas, not highways.
▶ KIMM: Autonomous driving runs smoothly even on rough dirt roads
Off-road environmental recognition technology, which can recognize foreign substances such as dust, mud, snow, and rain during off-road autonomous driving of construction and agricultural machinery and defense unmanned vehicles (UGV), and remove them in real time, has been developed for the first time in Korea. In the future, this is expected to be applied to industrial machinery such as excavators, dump trucks as well as autonomous vehicles in the defense field, and to be widely used in extreme environments where it is difficult for humans to operate.
The Korea Institute of Machinery & Materials(KIMM) develops “sensor protection modules,” “sensor signal correction technology,” “driving area recognition and driving control technology” that are applied to off-road unmanned vehicles in mountainous terrain, waterside areas, and snowy areas. We developed off-road environmental awareness technology and transferred the related technology to related companies.
[Nation]
▶ Anyang: Anyang City autonomous bus ‘Juyaro’ pilot operation
Anyang City’s two 18-seater autonomous buses, ‘Juyaro’, will begin trial operation with citizens starting on the 22nd.
At the level 3 of autonomous driving, it is equipped with sensor detection functions such as lane maintenance, automatic braking, and maintaining distance between vehicles, as well as autonomous cognitive functions such as lane change, stopping at a stop, and traffic signal recognition.
The Citizen can get on and off the bus at any bus stop on the route without a separate reservation. Infants under 6 years old are not permitted to board. During the trial run period, the bus is free of charge.
<< Other Countries >>
Currently, many industries are interested in or are introducing AI technology. And the automotive industry is preparing to introduce AI technology. Of course, there are differences in the approaches of automakers and startups. Automobile manufacturing companies that are focusing on developing technology for cars for sale are in the process of introducing technology to strengthen the Assist function. On the other hand, startups are preparing for more flexible(?) learning by moving away from traditional autonomous technology implementation methods and actively introducing data-based AI technology. Tesla is also attempting a technological leap once again by introducing this method.
In just six months, Wayve has unveiled an upgrade that applies LLM more actively to autonomous driving technology. Ultimately, it appears to be an effort to strengthen end-to-end autonomous driving technology. They are preparing AI learning in a more human-friendly direction by strengthening communication with people. Of course, it still appears to be in its early stages. However, the technology that companies need to prepare to implement fully autonomous driving technology in the future ultimately depends on how accurately and quickly they can recognize and perform communication with people who are passengers, not drivers. It can be said that concerns about such communication methods have already begun.
[Company]
▶ Tesla: Tesla slashes price for monthly Full Self-Driving subscription
Tesla is reducing the subscription fee for its Full Self-Driving (FSD) driver-assist software in a reflection of the company’s continuing financial hardships. The company slashed the price to $99 / month, down from $199 where it has been since at least 2021.
Tesla announced the price adjustment in a post on X, in which it described the Level 2 driver-assist system as “FSD (Supervised)” — a name tweak that’s meant to convey the fact that drivers are required to pay attention to the road and stand ready to take control of the vehicle. (The company has been criticized for failing to include proper driver monitoring and other protections against overreliance on the system.)
But FSD is not a perfect system. Some Tesla owners laud its capabilities, but others describe it as erratic and untrustworthy. The company has pushed out numerous software updates in a bid to improve its abilities, with the latest (v12) purporting to finally tap into what Musk calls “end-to-end neural nets.”
▶ AVIA: Autonomous Vehicle Industry Association Releases First-Ever “State of AV” Report
AVIA conducted a survey of its membership for an update on industry data and trends. Notably, the data reveals that AVs have driven nearly 70 million miles on public U.S. roads — equivalent to 293 round trips to the moon, or driving across Route 66 over 29,000 times. As the unified voice of the AV industry, this statistic applies exclusively to autonomous vehicles, which perform the entire driving task.
AVIA also asked CEOs of our member companies about their top priorities and potential obstacles for the future:
• Unanimously, CEOs characterized commercial partnership interest in the AV industry as “positive” or “very positive.”
• When asked about the biggest challenges facing their company, the most cited answer was “public policy challenges,” expressing a need for clarity of the rules governing AVs.
• Those surveyed agreed that policymakers need to increase their engagement with the AV industry.
• When asked about the “single most impactful” action for policymakers, top priorities from member CEOs included rules of the road from federal agencies, clear laws and regulations in the states, and federal legislation to encourage AV deployment.
Below are additional highlights from the survey findings:
• Top states for planned AV expansion include: Texas, New York, Arizona, Georgia, California, Illinois, Florida, New Mexico, Louisiana, Alabama, Mississippi, and South Carolina
• States with the best regulatory environments for AVs: Texas, Florida, Arizona
▶ Wayve: Driving with Language: Introducing Wayve’s Multimodal Driving Model LINGO-2
- Wayve’s LINGO-2 is a closed-loop driving model and the first vision-language-action model tested on public roads.
- This is the next step in Wayve’s pioneering work incorporating natural language to enhance the explainability of AI models.
- LINGO-2 unlocks new innovation for autonomous driving and marks a crucial step toward building confidence in the decision-making process of Wayve’s AI driving models.
[Nation]
▶ China: Hebei province opens first expressway section for intelligent connected vehicle tests
The Transport Department, Industrial and Information Technology Department, and Public Security Department of Hebei Province jointly announced the official opening of the G2 Beijing-Shanghai Expressway’s Hebei section in Langfang city (Jingjintang Expressway’s Hebei section) for intelligent connected vehicle road testing and demonstration applications.
The Jingjintang Expressway is the first expressway in China to support inter-provincial cooperative autonomous vehicle testing. The G2 Beijing-Shanghai Expressway’s Hebei section in Langfang is Hebei’s first highway section for intelligent connected vehicle testing. Its opening signifies the full availability of the Jingjintang Expressway for intelligen connected vehicle testing and demonstration applications, pioneering the cross-regional application of autonomous driving vehicles in China.
The complete opening of the Jingjintang Expressway signifies the integration of logistics transportation, including highway freight service, making “door-to-door” autonomous driving transportation a reality. This will greatly promote the exploration of C-V2X technologies along highways and accelerate the development of China’s intelligent connected vehicle industry.
[People]
▶ San Francisco: The divide over driverless taxis in San Francisco
How to reach to me
Please share your insights and opinions.
- Leave message at “Respond”
- Korean Website
Over the Vehicle !!!
