Countdown to 184 days.
One morning in April, an employee of Heduo Technology uploaded a photo in Feishu reminding that the day of delivery was getting closer and closer.
"One of the things we're under the most pressure right now is to complete the delivery of the GAC project." Ni Kai, founder and CEO of Heduo Technology, said in a recent interview with China Entrepreneur. Ni Kai worked at Microsoft, joined Baidu after returning to China, participated in the whole process of Baidu driverless from 0 to 1, and briefly took charge of intelligent driving in LeTV Auto. In 2017, he founded heduo technology, an autonomous driving company.
Source: Respondents
The first two years of the company's establishment have been moving forward in exploration: talking about cooperation with car companies, "10 car companies give you 10 different needs"; in 2019, the industry entered a trough, and many startups with bright backgrounds appeared on the death list. In order to achieve self-hematopoiesis, Heduo provides vehicle modification services with software and hardware customization to car companies. Until the beginning of 2020, the cooperation with GAC motor went smoothly, and Ni Kai had a bottom in mind about winning this big customer, and he felt that the light was in front of him.
In October 2021, GAC Group and Heduo Technology officially signed a cooperation agreement, announcing that the first Guangzhou automobile model equipped with Heduo's automatic driving system will be launched in 2022. In March 2022, GAC Capital, a subsidiary of GAC Group, completed its exclusive strategic investment in Heduo Technology, with an investment amount of hundreds of millions of yuan, accounting for more than 10% of the shares, becoming an important strategic shareholder of Heduo. Prior to this, the financing of Heduo Technology progressed to the B round, and the previous investors included NavInfo, IDG Capital, Sequoia Capital and so on.
Ni Kai believes that 2025 will be a key point in time. In addition to the new forces of independent research and development of autonomous driving, a large number of traditional car companies are breaking through, it is impossible for each car company to completely build its own team, build a system, and there will be no oligarchy of only a few car companies, and autonomous driving companies like Heduo will usher in opportunities.
At present, the mass production of high-level assisted driving is imminent, and Heduo Technology is about to accept market tests.
Niche entrepreneurial route
From the beginning, Heduo has determined a gradual route for autonomous driving.
In 2017, when Heduo was founded, the autonomous driving industry is catching up with a capital feast, and some entrepreneurs sigh that "you can raise funds with a resume". At that time, there were basically two routes, one was the L4 level unmanned route represented by Waymo, Baidu, Xiaoma Zhixing, etc. Another route is the progressive route, represented by Tesla, starting with L2 level assisted driving.
The former was considered to require greater software capabilities and was a path chosen by most entrepreneurs at the time. That path is favored by capital, and Waymo even valued $175 billion at one point in Morgan Stanley's 2018 assessment.
Ni graduated from the School of Computer Science of Tsinghua University and studied at the Georgia Institute of Technology in the United States, focusing on robotics and computer vision. At Baidu, Baidu mainly took the L4-level route, but Ni Kai eventually chose the progressive route as the direction of entrepreneurship. This has to do with his past experience.
While at Microsoft, Kai Ni was involved in the HoloLens augmented reality glasses project from a dozen people to three or four hundred people, and some functions such as three-dimensional reconstruction are very CPU-intensive and must be continuously optimized. This made Ni Kai realize that product launch is not as simple as doing research and issuing a paper, and the engineering link is crucial.
After returning to China to join Baidu, Ni Kai was responsible for cooperation with BMW. BMW's style is to think thoroughly before advancing, in this cooperation, Ni Kai saw a lot of problems that need to be solved to develop automatic driving. At that time, baidu's L4-level automatic driving solution "has too many pits in the landing", "even in the United States, the L4-level automatic driving program is in trial operation, and China's long-tail scenario is more." This made Ni Kai question the L4 path, and also prompted him to think about whether there is a path to embrace the automotive industry more deeply. From the data point of view, is the data brought by hundreds of L4 autonomous vehicles enough?
Ni Kai began to think about the gradual route of target mass production. At that time, he thought that LeEco Cars looked closer to his dream of realizing mass production of autonomous driving, so he decided to join. However, with the outbreak of the LeTV crisis, research and development is difficult to sustain. Ni Kai also thought clearly about the direction at this time, decided to achieve it through entrepreneurship, and officially established Heduo Technology in June 2017.
Source: Respondents
Although the general direction is set, it still needs a long process to get the mass production order of the car company, from contacting the car company, getting the PoC (Proof of Concept- proof of concept) project, and then determining cooperation and mass production.
Compared with other self-driving startups, Heduo's financing scale is not large. In order to achieve self-hematopoiesis, create products and engineering awareness, Heduo first launched the Xuanyuan platform three months after its establishment, providing automotive vehicle companies with automatic driving modification solutions, empowering "limbs (body) and eyes (sensors)", and customers developed "brains" on the platform. This model contributed most of Heduo's income in the early days, and at the same time, exercised Heduo's engineering ability on the vehicle platform.
But this is only a product positioned to "lay eggs along the way", in order to establish the industry position, Heduo needs to find a partner for mass production.
The direction is clear, but the specific implementation path is not clear. Ni Kai's self-evaluation is "a typical science and engineering man" who is not good at communication. But to promote the landing of technology, he must visit more car companies, local governments, and car companies are "almost the strongest party A in all industries."
What bothers Ni Kai the most is that talking about cooperation with 10 car companies can get 10 different needs. In 2018, the industry is not yet mature, the hardware cost has not come down, there is almost no solution to learn from, whether it is a car company or Heduo, there is a lack of deep enough understanding of the industry. In Ni Kai's words, this makes it difficult for both sides to fully convince each other when talking about cooperation, "like two primary school students discussing.".
Later, with the application and evolution of Tesla's assisted driving Autopilot, the market was gradually educated, and the two partners could discuss more symbolic content. In 2018, Heduo launched two product prototypes of highway autonomous driving and intelligent parking, and began large-scale testing in 2019 to push the prototype products to maturity.
In this process, stepping on the pit and adjusting is inevitable. For example, the intelligent parking function, at that time, Heduo found that only the bicycle intelligence could not achieve this function after cooperating with the car company, so it proposed not only to transform the vehicle, but also to transform the parking lot, and arrange sensors and calculation units in the parking lot. But in fact, the parking lot is not willing to spend 200,000 to 300,000 yuan to install equipment. The project was temporarily shelved.
In addition, it is common to be rejected by customers due to price, technical routes and other reasons. "I think this business model of to B is to constantly consider what customers need, and today this project was rejected, and the next project may be better by reducing prices or making technical and product improvements." Ni Kai said.
Key customers
In the second half of 2019, GAC began to contact Heduo. Yuan Feng, general manager of GAC Capital, told China Entrepreneur that at that time, GAC had done comprehensive research and contact with powerful autonomous driving startups in the market, and finally Heduo Technology passed the selection criteria of GAC Complex because of its engineering capabilities.
Yuan Feng pointed out that the requirements for engineering capabilities in front-loading mass production are very high, "the whole set of equipment needs to match the customized needs of car companies, can be loaded into the car, and at the same time ensure the normal operation of a long period of time after that." "Some of the core personnel in the Heduo team are from Bosch, which improves the mass production capacity of Heduo. In addition, during the long test process, Heduo showed a high degree of cooperation.
Woduo, who is underfunded, urgently needs to prove his ability through this project. At the end of 2019, the autonomous driving industry has almost cooled to a freezing point, and startups have fallen out of favor with capital, and they are worried about whether they can survive the winter. In early 2020, the epidemic struck. For nearly a year and a half, Heduo did not have new financing. On the one hand, Ni Kai contacts as many investors as possible, on the other hand, streamlines personnel and opens up and reduces expenditure.
With the advancement of cooperation with GAC, Heduo, which has business endorsement, announced the completion of two rounds of financing, A+ and B rounds in August 2020 and June 2021. In October 2021, GAC Group signed a cooperation agreement with Heduo, and in March 2022, GAC Capital strategically invested in Heduo.
After alleviating the financial pressure, the smooth delivery became the biggest challenge for Heduo. Ni Kai said that first of all, this is a product with very high requirements for engineering capabilities and AI technology capabilities, in addition, this is the collision of two systems of car companies and technology companies, and the research and development and cultural systems of both sides are different and need to be constantly aligned. Heduo set up an office near GAC and held meetings with GAC every day.
In terms of team, Heduo Technology is also recruiting.
Jiang Jingfang. Source: Respondents
Jiang Jingfang worked at Bosch for 21 years, serving as the head of the ADAS business unit in China, senior vice president, etc., and announced in June 2021 that she joined Heduo as a senior vice president and partner.
In recent years, Jiang Jingfang began to think about a question, Bosch has occupied a pivotal position in the field of traditional automobile supply chain, and under the new intelligent trend, car companies have emerged in an endless stream of new demand for automatic driving, and higher requirements have been put forward for demand matching and R& D efficiency.
Jiang Jingfang received "olive branches" from several startups, and she eventually chose to join Heduo because she believes that Heduo is "very focused" and has always aimed at mass production. At Heduo, she will be directly responsible for more businesses including quality and supply chain, and has more room to play than at Bosch. Ni Kai's sincerity also touched her. Several times I received WeChat from Ni Kai at two or three o'clock in the morning, and once he rushed to Suzhou on the weekend to dine and chat together. I have other options, but the boss who ultimately wants to work with must be something you can trust, rely on, and cooperate with. Jiang Jingfang told China Entrepreneur.
After joining Heduo, Jiang Jingfang needed to land a product from 0 to 1, going through the entire chain from demand, project management to functional development and landing, and the new system was more complex than when she was at Bosch. Intuitively, the number of radars and cameras is more, and the integration with high-definition maps requires more different partners and previous different partners, and there is more synergy with car companies to jointly define the needs. "We often joke that this is a pilot project, and we can't do it without taking off a layer of skin." Jiang Jingfang said.
Ni Kai also knew in his heart that it was impossible to have a smooth sailing. In the face of employees' questions about the delay of the project, Ni Kai wrote in the company's Feishu circle: "No company's first-generation products are smooth sailing, without being pressed to the ground by customers and rubbed, mass production is impossible." Yes, I will also be anxious from time to time, but more than that, I still feel the determination and motivation to do this well. ”
Match point is approaching
In addition to going all out to promote the GAC project, cooperation with other car companies is also advancing. A problem that cannot be ignored is that data and software have become the core competitiveness of autonomous driving, and the acquisition of data depends on car companies, which makes autonomous driving companies need to consider multi-party relationships when cooperating with multiple car companies.
Jiang Jingfang said that the mainstream of the industry is no longer the traditional OEM-supplier cooperation model, but through the combination of mass production orders + investment, forming a deeper cooperation model. Heduo and GAC are carrying out in-depth communication and cooperation from the R&D needs, code, and underlying software of autonomous driving, and gradually promoting the mass production of high-end automatic driving solutions.
Yuan Feng said that after the cooperation between the two sides to jointly develop the products, through the iterative upgrading of GAC's data access drive product technology, the customer experience will be continuously improved to achieve GAC's leading level in the development and application of autonomous driving technology. When the relevant products and technologies are verified by the market, they can gradually serve other car companies and further polish the products better through larger-scale data access. He further explained that a similar logic has been verified in the cooperation between GAC and China New Air (formerly AVIC Lithium Battery). For a period of time after GAC Capital Invested in Innovation Airlines, most of the power lithium battery products of China Innovation Airlines were supplied to GAC, and after the products were verified and matured in the market, they were gradually opened to other car companies.
From the perspective of the industry environment, the mass production of autonomous driving has reached the race point of competition delivery, all companies are racing against time, and the first batch of paperrollers are about to be tested by the market.
On the assisted driving track, Tesla is in the lead, and Wei Xiaoli is also in the layout. In February 2021, Tesla re-launched the enhanced automatic driving function (EAP) optional package in the Chinese market; in the first half of this year, Xiaopeng Motors plans to launch a city-wide NGP intelligent driving assistance function; Weilai said that it is expected to achieve assisted driving on the NAD system equipped with models such as ET7 and ET5 this year; and the ideal car L9 is also equipped with the intelligent driving system "Ideal AD Max".
Technology companies and car companies to accelerate cooperation, momenta and SAIC cooperation of the first model Zhiji L7 is expected to start delivery in May this year, can provide parking, high-speed, urban and other scenarios of automatic driving assistance function; the end of the year zhixing is expected to be assisted driving system (including high-speed NOH, urban NOH, etc.) landed on 34 models of Great Wall Motors.
Ni Kai believes that by 2025, the autonomous driving industry will give birth to 2 to 3 head companies that provide mass production solutions, before that, Heduo needs to think about how to win orders and complete millions of deliveries.
Source: Respondents
As for profitability, Ni Kai said that it is expected to achieve profitability in the next 3 years and will also be committed to completing the target ahead of schedule.
Before starting a business, Ni Kai admired Mobileye, "It only makes a minimizable implementable product (MVP) and achieves mass production to the extreme." After starting a business, Tesla had a greater impact on Ni Kai, "closing the data loop to the extreme, whether it is a chip or an automatic driving system FSD, it is ultimately for the data closed-loop service."
In the 5 years since its establishment, the company has grown to more than 400 people, and Ni Kai is also making up for his shortcomings and improving his communication and management skills. Before the Reading Day on April 23, he recommended two books to employees, one was an interview with Zuo Hui, the founder of Shell House and Chain Home, and the other was "Huawei Team Work Method". Ni Kai lamented zuo hui's phrase "do difficult and right things", and Huawei is "a company that is very worthy of study and learning".