Written by Tang Rui (Founder and CEO of Zongmu Technology & Silkworm Robot)
Editor / Qian Yaguang
Design / Zhao Haoran
In the highly competitive field of intelligent driving, differentiation through innovation and creating more value for car manufacturers have become the key to the survival of intelligent driving startups.
On June 15, Tang Rui, founder and CEO of Zongmu Technology & Silkworm Robot, delivered a keynote speech entitled "Drop'n Go, How to Be an Ecological Partner of OEMs in the Era of Intelligent Driving" at the 16th China Automotive Blue Book Forum, explaining to us how intelligent driving companies refuse to roll up and use their imagination.
In his speech, he said, "Today's AI needs to help us replace some of our Tasks and make them easier, not replace some of our Jobs." He believes that it is possible and probable that L4 may be the first to realize the scenario of passenger cars in closed parks.
Zongmu's second-generation Drop'nGo technology (low-speed autonomous driving) based on a surround-view camera and 4D imaging millimeter-wave radar can also be used to valet parking even when it is raining or the camera is covered.
Tang Rui's other entrepreneurial project, the silkworm robot, is to solve the problem of mobile charging of cars in parking lots, and now mobile charging robots can provide charging services for any car.
Drop'n Go and mobile charging robots connect a series of actions such as automatic parking and charging in all parking lots, and provide a network for deeply penetrated first- and second-tier commercial real estate, which can continue to create a better user experience and higher value.
The following is a transcript of Tang Rui's speech, which has been deleted.
Thank you very much for the opportunity to share some of our thoughts with all of you in the automotive industry. Today's theme is "imagination", at the beginning, President Jia was also clearly opposed to involution, and the previous Mr. Yu, Mr. You, and Mr. Qin were more likely to discuss how to refuse involution and make value in today's era of involution from the perspective of car companies.
Zongmu Technology has always been a label in the domestic intelligent driving industry, today I have two identities: as the founder of Zongmu Technology, and I am also the founder and CEO of the new business - silkworm robot.
For a long time, there has been a strong perception in China that Zongmu is doing a good job in low-speed automatic driving and parking. At the press conference when the M5 was just launched, Mr. Yu was still praising our parking system for doing a good job. We have also been building a brand of low-speed autonomous driving, Drop'n Go, so today I have a small topic called "Drop'n Go, how do we be an ecological partner of OEMs in the era of intelligent driving". Because today is a highly involuted environment, there are a lot of discussions and analyses in the industry, how can we create value? How to avoid the zero-sum game between intelligent driving companies and OEMs, because this is a very unhealthy state in the industry.
Why do we call the Drop'n Go? Because we want to push L4 to the point of mass production. Today, the level of intelligence in the industry is gradually improving, and the mainland has also issued L3 test licenses, and L4 has always been the highest goal of the entire industry.
The most important crown jewel in the industry's pursuit over the past few years has been L4 Robotaxi, but how far is it from being realized? Now it's actually very controversial, some people think it's very close, and there are indeed commercial trial operations in various places, but I think there are also a lot of challenges. Tesla's Elon Musk recently announced on Twitter that Robotaxi will be launched on August 8, and many people are waiting to see if it can really usher in L4 Robotaxi on a large scale on August 8?
A month ago, Auto Business Review published a well-translated, interesting compilation of an article that is worth sharing: Tesla's Robotaxi Business is at least four years behind Waymo. When Tesla started hiring remote operators, we knew they were serious about Robotaxi.
Personally, I have a relatively conservative view on whether L4's Robotaxi can be truly commercialized on a large scale soon, because I think it is not only a technical issue, but also a lot of legal, legal and social issues. Today, many times ride-hailing drivers are the most important safe haven for the unemployed before they find their next job, and it is hard to imagine that Robotaxi will take such jobs away from society after it is truly commercialized on a large scale.
I feel that today's AI needs to help us replace some of our Tasks and make them easier, not replace some of our Jobs. This point of view is also borrowed from what Li Feifei said at a previous AI forum, what AI should help is to make our work, life and tasks simpler, rather than replace our jobs.
The landing of L4 has always been the goal pursued by the entire automotive industry, and our view is that it is possible and probable that L4 may be the first to achieve the scenario of passenger cars in closed parks.
Because compared with high-speed, urban and closed parks, closed parks have two characteristics, which are most likely to be converted to L4: first, closed parks are indeed unlikely to have fatal accidents, because fatal accidents are prone to occur in high-speed and urban areas, and it is difficult for OEMs to have a strong willingness to upgrade from L2 to L3 or even L4, which means that they have to take the potential fatal risk on their shoulders. Generally, many OEMs are conservative in their view, and the speed of closed parks is relatively low, and they are less prone to accidents.
The second point is that when we arrive at the closed park, we have already reached the final point of the journey, which is different from highways and urban areas, and it is impossible for the people behind the wheel to be released. If you spend a long time looking for a parking space in a closed park, it is better to drop off and go away from the car and go to your destination, whether it is to work, work, watch a movie, eat, or go to entertainment. We call the technology protocol stack of the L4 closed campus Drop'n Go, which is also the direction we are tirelessly promoting and progressing.
The biggest leap from L2 to L4 is not in technology, but in business model. L2 sells a product in many cases, it may be the intelligent driving system sold to the OEM, or the OEM sells a car to a passenger car customer, L2 is basically the business model of the product.
However, once in the case of L4, the responsibility is transferred from the user of the product, whether it is the Robotaxi model or the autonomous valet parking mode in the park, it is necessary to pass through continuous service, because only through continuous service can it be transferred in a real sense, in order to be responsible not only for the car owner, but also for the traffic road participants.
To give two examples, it turns out that some companies need remote dispatchers when they want to make Robotaxi, and there is no way to deal with very complex game scenarios, how can these cars be removed through remote driving? In fact, on the open road, this remote operation can sometimes be very inefficient. I once read a report that in San Francisco, Waymo's car suddenly got stuck in the middle of the road, was the remote system able to move the car in a timely and efficient manner? It's actually a painful thing, so L4 may not be as fast.
Similarly, in parking lots, whether today's AVP can achieve L4 in the real sense is also strongly dependent on O&M, because without O&M help in such extreme situations to solve problems such as moving cars and getting out of trouble, AVP also has no way to carry out large-scale mass production. This is the result of our thinking over the past few years.
As early as 2019, we were thinking about solving the two most complex service products in AVP landing scenarios: without O&M, AVP would not be so fast, even if it would be much faster than Robotaxi on the public road.
From the beginning of our first generation of Drop'n Go in 2017 to today, many car companies have released independent valet parking products, but we can really be confident that we can buy a car in a very short period of time and be able to use the AVP function in 80%-90% of the parking lots in Beijing, Shanghai, Guangzhou, and Shenzhen? I don't think the day was that fast. We need to solve two problems: first, to achieve the transfer of responsibilities through operation and maintenance; Second, it is possible to generate a map of mixed parking lots efficiently and cost-effectively.
Combined with the fierce competition in today's industry, I have made some strategic adjustments to the development of the enterprise and the thinking of the entire industry. As I said, we released the first generation of Drop'n Go technology in 2017, and in the underground parking lot of Beijing, we have demonstrated that we can achieve autonomous valet parking or remote summoning in a well-lit parking lot with AI neural network technology based on surround view cameras.
In 2019, we released the second generation of Drop'nGo technology based on surround-view cameras and 4D imaging millimeter-wave radar, which can also achieve AVP when it is raining or the camera is covered. Earlier this year, we released the next-generation Drop'nGo technology of BEV Transformer + OccupancyNet. In the process of technology development, in addition to the iteration of technology, I think more about how to further upgrade and evolve the Drop'nGo service.
What we want to create is an L4 experience, which means that you can Drop Off and then Go Away, and the rest should be worry-free and non-sensitive for consumers. Autonomous valet parking is one of the most important functions, but Drop'nGo's own connotation and extension can be further extended. From the connotation of Drop Off, Go Away, and letting the car park by itself, it has evolved to provide parking services for many car owners, including parking space grabbing, automatic location calculation, mobile charging and other aftermarket value-added services. In this way, we hope to avoid direct head-to-head competition with car companies in the intelligent driving industry. Today, many people are anxious, is there room for independent intelligent driving companies? There are a lot of articles in the industry that are anxious for us, in fact, more often, we should think about how to become complementary to car companies in the real sense, rather than a simple value game. Therefore, we want to do something different and provide value for car companies by continuously exploring the ecology and services of parking lots.
Based on this, at the end of 2022, we founded the Silkworm Cong Robot, the earliest thinking is to solve the problem of mobile charging of cars in parking lots.
At the end of 2023, we conducted a commercial trial operation and achieved quite good results. Our energy robots can not only charge new energy vehicles, but also store energy for parks. In the end, we found that the essence of this matter is that it can provide a more effective combination of autonomous driving + energy storage technology in the field of distributed energy structure today to form a new type of energy or power transmission and distribution, and its applications include charging new energy vehicles, industrial and commercial energy storage for parks, and many temporary and emergency electricity consumption.
On January 18 this year, we held a relatively large mobile energy Internet innovation forum and product launch, and invited many government and industry leaders from France, Hungary and Singapore to witness the product launch with us. In April this year, our products officially rolled off the assembly line for mass production. During the "May Day" period, the first large-scale mobile charging carnival in human history was held, and about 100 energy robots, Lightning Treasure, were deployed to provide charging services for tens of thousands of car owners in Shanghai Disney Resort.
On May 22nd, we went out of the country to Europe and North America. Some of these pictures are in North America, some are in Hungary, we participated in the Hungarian GreenTech exhibition, and the Hungarian energy minister said that our lightning treasure is the most innovative product form in the whole GreenTech exhibition.
From Zongmu to Silkworm Cong, what we hope to do is to truly become an ecological partner of OEMs in the era of intelligent driving. After going offline in April this year, new energy charging robots have been deployed in many parks in Shanghai in May. In this process, we will provide charging services, map services, etc., and next year we will aim to achieve mobile charging services for 5,000 large parking plants in first-tier cities.
Zongmu and Silkworm Cong are actually one company, we have the business of cooperating with the OEM, and hope to continue to be able to empower the OEM as an intelligent driving system, whether it is a parking system or a driving and parking integrated system, but most importantly, we hope to provide Drop Off, Go Away SDK, which can provide a large number of value-added services for car owners in the parking lot, including On Demand mobile charging, motor parking spaces, car finding, reverse car search, AR navigation, etc. We will map all the major large-scale commercial centers, office parks, and above-ground and underground maps in first-tier cities, and use this digital map to do value-added business, which will derive many new business models.
Silkworm Grove Robot is really doing two things: it can provide charging services for new energy vehicle owners through the Silkworm Cluster Robot itself, and at the same time, when deploying comprehensive solutions, it will also deploy a lot of high-power two-way energy gateway FlashBase, in addition to providing two-way charging and discharging to our Silkworm Grove Energy Robot, it can also provide up to 100KW two-way V2G service for future new energy vehicles.
These things can allow us to connect a series of actions such as automatic parking and charging in all parking lots, and provide a network for deeply penetrating first- and second-tier commercial real estate, which can continue to create a better user experience and higher value.
In conclusion, I would like to say that I very much agree with Dr. Jia Ke's words, this era needs a little imagination, we need to refuse to roll in, and we need to go out hand in hand. Not only the intelligent driving business, but also the mobile charging business, we really hope to become a very strategic cooperative relationship with OEMs. As I mentioned earlier, we have been penetrating many markets around the world since April this year. We have also noticed that in the process of new energy going overseas, the biggest pain point encountered in many cases is the imperfection of the energy replenishment network.
I would like to conclude my presentation with a few short stories: the picture shows a new energy Mercedes-Benz truck that specializes in rescue services at the GreenTech exhibition in Hungary, but this rescue vehicle is out of power and needs our FlashBot to come to the rescue.
Another photo is that we are in the process of charging, there is a local Kia agent who can't find a charging pile everywhere, and after using our service, he said okay, because a very important reason why electric vehicles can't be sold is that there is no way to charge them quickly. I took a photo and sent it to Mr. Shen Fei of NIO, and they said that they met your car here, and they were also surprised.
In fact, whether it is an OEM or an intelligent driving ecosystem, everyone is indeed facing a lot of pressure. At such a point in time, we should use our imagination as President Jia called for, and refuse to simply roll up some low-cost, worthless, and incapable of bringing positive value to consumers and the industry, but should think about how to go out in a real sense and create better things for the world.