At the China Artificial Intelligence Conference held on July 21st, Professor Li Deyi, an academician of the Chinese Academy of Engineering and chairman of the China Artificial Intelligence Society, raised a thought-provoking question: “If we can't even achieve automatic parking, how can we reach Level 3 autonomous driving?†This statement sparked discussions about the current state and future of self-driving technology.
The concept of automatic parking has been around for over a decade. In 2005, Citroën introduced its first automatic parking assist system, known as the City Park system, which only took control of the steering wheel. Today, such systems have evolved significantly, helping both novice and female drivers park with ease. However, despite these advancements, the transition from manual to automated driving remains challenging.
According to the Society of Automotive Engineers (SAE), autonomous driving is divided into five levels: L1 (driver assistance), L2 (partial automation), L3 (conditional automation), L4 (high automation), and L5 (full automation). While many vehicles on the road today, including Tesla models, are classified as L2, they still require human oversight. The 2016 Autopilot Report by the California Department of Motor Vehicles noted that the average intervention frequency was as low as 0.2 per thousand miles, highlighting the limitations of current systems.
Moving from L2 to L3 involves a critical shift—transferring control from the driver to the vehicle. This transition presents several challenges, as identified by Li Deyi: How do we determine when to hand over control? How do we measure the success of this handover? And most importantly, how do we handle accidents that may occur during the process?
Audi’s upcoming L3-certified A8 model is expected to be the world’s first commercially available fully autonomous vehicle. But the question remains: Will traffic authorities issue licenses for such cars? And will ordinary drivers feel confident using them without a traditional license?
Li Deyi also highlighted the challenges faced by L3 technology in Germany and the broader industry. The key to breaking through the L2 ceiling lies in artificial intelligence, particularly in improving the car's ability to perceive, reason, and act independently.
He emphasized that once certain conditions—such as geofences, weather limits, or human behavior constraints—are breached, the system must immediately hand over control. These transitions can be riskier than manual driving, making safety a top priority.
The core question of L3 is whether the challenge lies in improving the vehicle itself or in redefining the role of the driver. The "car" refers to a machine equipped with advanced software capable of autonomous driving, while the "human" represents the replacement of driver cognition with robotic intelligence. This includes memory, decision-making, and behavioral skills.
During autonomous driving, the driver’s role in predicting and controlling the environment remains irreplaceable. This led to the development of the "driving brain"—a system that combines memory, computation, and interaction to create a more intelligent and responsive vehicle.
Unlike traditional sensors like radar, the driving brain must process complex cognitive tasks, making it a crucial component in the smart car ecosystem. In-vehicle computers and robotic operating systems cannot fully replace this function.
Li Deyi’s team has collaborated with companies such as Yutong, Chery, and SAIC to develop and commercialize the driving brain through real-world testing. Their work highlights the importance of integrating microelectronics and AI to build safer, smarter vehicles.
In summary, the journey from L2 to L3 autonomous driving is not just a technical challenge but also a societal one. It requires not only advancements in hardware and software but also a rethinking of how humans and machines interact on the road. As the industry moves forward, the focus will remain on ensuring safety, reliability, and public trust in autonomous technology.
Types of Intel Motherboard
Consumer Motherboard: These are designed for home and office users, offering a balance of performance and affordability. Consumer motherboards come in a variety of form factors, including ATX, Micro-ATX, and Mini-ITX, to fit different case sizes and usage scenarios.
Gaming Motherboard: Built for gamers, these motherboards offer enhanced features like better audio quality, more robust power delivery, and advanced cooling solutions. They also often come with RGB lighting and other aesthetic enhancements to make your gaming rig stand out.
Workstation Motherboard: These are designed for professional users who need high-performance computing capabilities. Workstation motherboards typically offer more expandability options, support for multiple processors, and advanced error-correction features to ensure reliable operation.
Server Motherboard: Used in data centers and enterprise environments, server motherboards are designed for maximum reliability and scalability. They often support multiple CPUs, large amounts of memory, and redundant power supplies to ensure uninterrupted operation.
gaming motherboard,mini itx motherboard,motherboard pc,itx motherboard
Shenzhen Innovative Cloud Computer Co., Ltd. , https://www.xcypc.com