Q: What was the reason behind the recent organizational restructuring of the autonomous driving team?
A: Shaoqing:
The shift to an end-to-end approach was driven by the previous structure, which consisted of multiple modules like perception, control, and prediction. Each module had its own team working on models and rules. Now, the focus is on integrating everything into a unified end-to-end approach to enhance the user experience. Previously, the process was sequential; now, it’s more integrated. The change in technical direction necessitated an organizational change.
Q: Is there confusion in NIO’s autonomous driving approach, given that the main brand uses multiple sensors while LeDao (ONVO) uses a pure vision approach?
A:
The more sensors, the better, and it’s nonsensical to say LiDAR is useless. However, LiDAR’s value varies at different stages. It’s similar to airbags; more airbags are generally better for safety. The choice of using LiDAR or not is a business decision. For NIO, LiDAR is standard on some models, like the ET9, which has three LiDARs. Whether a car has LiDAR depends on the cost, and it’s unrealistic to expect a 100,000 RMB car to have the same features as an 800,000 RMB car. We are confident that LeDao’s pure vision-based assistance system will perform well. Currently, both brands’ systems are developed by the same team.
Q: What are the reasons BBA (Benz, BMW, Audi) customers choose NIO?
A:
Most NIO customers come from traditional high-end gasoline vehicles. They choose NIO for several reasons:
1. No range anxiety: NIO offers a pure electric vehicle experience with a charging and battery swapping network, providing easy access to charging services.
2. Intelligent features: Traditional cars often correlate intelligence with price, but NIO offers advanced smart features that are hard to leave behind once experienced.
3. High-end brand community: NIO maintains a premium brand image with aesthetics, culture, and attitude, making high-end gasoline car owners feel they are not downgrading.
High-end brands must excel in various aspects, and NIO is positioned as the next choice for BBA customers.
Q: How will chip restrictions affect NIO?
A:
U.S. chip restrictions have impacted China’s automotive industry. Since last October, access to the world’s best cloud training chips has been limited. This highlights the importance of collective intelligence, which considers both cloud and on-device computing power. While the current situation for on-device inference chips is manageable, future developments are uncertain, and we must actively adapt to changes.
Q: How did SKY OS come about, and are there any behind-the-scenes stories?
A:
NIO began developing its own technology in 2015, focusing on foundational software. The earliest work was related to safety, based in Silicon Valley, and achieved recognition at a North American hacking competition. Building foundational systems requires careful planning, not just research for the sake of research. Despite difficulties in 2019, we committed to developing an operating system after recovering somewhat in 2020. This effort included developing middleware and accumulating relevant knowledge. NIO has since recruited a top global team, specializing in microkernel systems rather than application software. This team is dedicated to foundational work, and as we move forward, SKY OS will increasingly deliver value and efficiency. The operating system development is a strategic move, not just for the sake of R&D.
Let’s take a look at the contents of the NIO group interview in 2024:
Interviewees: Li Bin, Qin Lihong, Wang Qiyan, Ren Shaoqing (mostly answered by Li Bin)
1. Investment in SkyOS and Chip Development
• It’s not typical for companies to invest this way; most focus on components, but smart car companies should adopt a layered and decoupled approach. NIO has pursued this approach aggressively, decoupling the battery pack from the vehicle. If the high-voltage platform stabilizes, the same battery pack can be used for four or five generations of vehicles. Initially, it might seem like a lot of investment, but it gets easier over time. Building a solid foundation is crucial, like laying a foundation for a 100-story building; it takes time but speeds up once above ground.
2. Comparison of NIO’s System with Xiaomi and Huawei
• NIO’s system is car-centered, differing from Xiaomi and Huawei’s approaches. Each company has its unique strengths. There are valuable lessons to learn from HarmonyOS.
3. Will NIO Share Its System?
• Current operating systems are fragmented, inefficient. NIO’s native system offers high performance for car scenarios. We are open to sharing it with the industry eventually, focusing first on internal users, then component partners, and finally, whole vehicle partners if interested, similar to our battery swap model.
4. Adjustments in the Autonomous Driving Team
• Organizational structure follows technical architecture.
5. NIO’s Approach to Intelligent Driving
• More reliable sensors and stronger computing power are beneficial. Dismissing LiDAR is shortsighted. Different models serve different purposes at various stages. NIO believes in using LiDAR and pure vision, with confidence in their capabilities. It’s unrealistic to expect low-cost cars to have high-end configurations.
6. Value Creation and High Investments
• NIO has not neglected features and configurations that users highly perceive, despite significant long-term investments. The company balances visible innovations with foundational technology development.
7. “Moat” in Intelligent Driving
• The journey is long; even in battery swapping, it’s too early to claim a moat. In a highly competitive and rapidly evolving market like China, companies must focus on foundational models.
8. Standards for Evaluating a Good Car
• Most NIO users come from traditional high-end fuel cars, choosing NIO for reasons like range anxiety-free electric driving and a high-end brand community. NIO aims to capture a significant portion of traditional BBA car users switching to new energy vehicles.
9. Impact of No New Models This Year and Brand Day Investments
• NIO’s planning is long-term. Despite not launching new models this year, the company has maintained a steady monthly sales rate of 20,000 since May. NIO doesn’t authorize weekly sales reports.
10. Impact of U.S. Chip Policy on China’s NEV Industry
• U.S. chip restrictions have impacted China’s auto industry, especially in high-computing chips, although the impact on edge inference is manageable.
11. SkyOS Development and Relationship with AutoSAR
• NIO started developing its system in 2015, with significant focus on foundational software. The decision to create their own system stemmed from issues with AutoSAR and the need for a long-term technical roadmap.
12. Robots and Robotaxi
• NIO is not interested in Robotaxi, seeing autonomous driving’s value in enhancing safety and reducing driver fatigue, rather than eliminating jobs.
13. Will NIO Make Home Products with NOMI?
• No current plans, as the focus remains on cars.
14. NIO’s Standing in the Industry
• Many are unaware of the real NIO. The company emphasizes its responsibilities to shareholders and users, with a commitment to continuous improvement and innovation.
15. NIO Phone Plans
• NIO aims to offer a flagship phone focused on software experience, not competing in large scale. The company sees potential in providing a clean phone experience, especially to non-NIO car users.
16. Time Required for Sufficient Training Intensity
• The goal is to improve user experience and R&D efficiency. The team size will not decrease.
17. Concerns About R&D Investment
• NIO invests approximately 3 billion RMB per quarter in R&D, around 10 billion RMB annually, ensuring the company remains competitive in the final rounds of the race.
