With 5.6 billion yuan of funds pouring in in a single month, the autonomous driving track is still searching for answers to the future

Autonomous driving has ushered in a new round of "battle among the gods".

From May to June this year, a total of 16 domestic companies in the field of autonomous driving received financing, with a total financing amount of more than 5.6 billion yuan. New car manufacturers such as Xiaomi and major investors with a keen sense of smell have joined this carnival, and the commercialization of autonomous driving seems to be getting closer and closer to us.

On June 23, WeRide, the earliest self-driving company in China, received a C round of financing with a total amount of US$310 million. Investors include Renault-Nissan-Mitsubishi Alliance Fund and other well-known institutions. So far this year, WeRide has completed more than US$600 million in financing.

This trend is still continuing.

On July 16, Juefei Technology, which focuses on vehicle-road collaborative technology, announced the completion of its A+ round of financing, led by a number of strategic investment institutions, with old shareholders such as Hegao Capital also continuing to increase their investments. The amount of financing reached tens of millions of yuan. This is also the second round of financing completed by this young company established at the end of 2018 within two months.

Those companies that have not been favored by capital are not willing to lag behind, and have immersed themselves in research and development, waiting for their own opportunities: Zhixing Technology announced in May that it had begun researching high-level autonomous driving systems with Mobileye, and is expected to deliver products with this function to domestic first-tier OEMs in the fall; AutoX has obtained driverless licenses and demonstration operation qualifications in many cities in the United States and China this year, and achieved China's first fully driverless driving at normal speeds. Its fifth-generation driverless system was also officially announced in July.

The momentum of new energy vehicles is leading the take-off of the entire domestic industrial chain, and autonomous driving is undoubtedly the most complex and important part of it.

A complete autonomous driving system is mainly divided into four parts: perception system, decision-making system, execution system, and communication system. It involves hardware such as radar, sensors, high-precision maps, chips, and also requires a large amount of data as a basis.

Therefore, powerful automakers will try every means to master a complete set of autonomous driving systems. SAIC calls autonomous driving the "soul" of new energy vehicles and explicitly rejected the full-stack solution provided by Huawei.

However, no matter which model car companies prefer, it is an indisputable fact that the development of autonomous driving needs more promotion from third-party technology companies. For this reason, many third-party companies focusing on autonomous driving technology have sprung up like mushrooms after rain. They have different development focuses and operation strategies, and explore the future of autonomous driving technology in their own ways.

Juefei Technology: Bravely walking on the "single-plank bridge" of vehicle-road collaboration

At present, the autonomous driving industry has two main development directions: single-vehicle intelligence and vehicle-road collaboration. In theory, the latter is a more ideal form: vehicle-road collaboration can reduce the requirements for single-vehicle hardware and save costs in the long run. More importantly, no matter how complete the sensing devices on the vehicle are and how advanced the algorithm is, detection blind spots are always unavoidable. In the event of special circumstances such as network disconnection, the consequences are even more disastrous. If the road can also become intelligent and form a "road command vehicle" situation, the accuracy and safety of autonomous driving can be qualitatively improved.

However, vehicle-road collaboration not only has extremely high requirements for V2X technology, but also requires that many devices on the road have Internet perception capabilities. This requires the concerted efforts of industry participants and even government departments. The initial costs are visibly high, and it has become the most difficult path in the field of autonomous driving.

But the difficulty also means that there are relatively few competitors. Juefei Technology, founded at the end of 2018, has embarked on this "single-plank bridge". Juefei Technology insists on the technical route with "multi-sensor fusion" as the core, and empowers the vehicle and road ends through the "technology + data" model.

For vehicles, Juefei Technology provides high-precision map data and fusion positioning algorithm services. For roadsides, they can provide digital road fusion perception solutions, providing high-precision fusion perception and positioning algorithms for roadsides, regional high-precision road data services, dynamic traffic information and other services. At the same time, Juefei Technology encapsulates core algorithms and technologies so that it can be quickly customized for different customers, which not only ensures the flexibility of solution adaptation, but also avoids additional costs.

At the 4th World Artificial Intelligence Conference held on July 8 this year, Juefei Technology released the roadside fusion perception product and solution Zhihuan System.

The Zhihuan system has built-in laser radar, millimeter-wave radar, industrial cameras and other equipment in the roadside monitoring equipment. The data collected by these devices are recorded and digitally reconstructed through a high-performance computing platform, and then analyzed and processed into road condition data that can be directly used through edge processing algorithms, and finally sent to the car. At the same time, the data of the Zhihuan system will also be uploaded to the cloud synchronously and sent to nearby vehicles for them to adjust at any time.

The Zhihuan system integrates laser radar, industrial camera, millimeter wave radar and high-computing computing unit into the same body, making the roadside observation system high-precision, fully digital and full-life cycle. When combined with Juefei Technology's rich scalable multi-sensor fusion traffic algorithm library, this system can further reduce computing latency, perceive vehicles more accurately and control them in a timely manner, and provide a sensor, software and hardware integrated solution for various application scenarios of vehicle-road collaboration.

Juefei Technology's integrated solution makes vehicle-road collaboration no longer a seemingly illusory concept, but a real possibility of implementation. Juefei CEO Li Dongmin said: "With the advent of China's smart transportation and vehicle-road collaboration era, integrated perception technology will become the core key technology of vehicle-road collaboration, and the development of roadside intelligence will also usher in an unprecedentedly broad market space."

WeRide: A pioneer in the industry

In the recent wave of financing for autonomous driving, WeRide has the highest amount of financing. This is not surprising. Founded in 2017, WeRide is the first company in China to implement autonomous driving, and it also has the industry-leading L4 autonomous driving technology: In February this year, the California Department of Transportation released the 2020 full-year autonomous driving data. In terms of MPD (miles per intervention) ranking, WeRide ranked sixth in the world.

Han Xu, the founder of WeRide, believes that in the autonomous driving industry, general technology platform companies have the ability to launch a "dimensionality reduction attack" against their peers. Therefore, WeRide has always focused on L4 autonomous driving technology and built a high-dimensional platform-based autonomous driving.

WeRide provides a full-stack software and hardware solution, with core technologies such as self-built centimeter-level high-precision maps, fast positioning, and unmanned driving algorithms, as well as 360-degree full-coverage sensor modules, front-mounted laser radar and other hardware, which can achieve detection within 250 meters. At the same time, using multi-sensor cross-validation, WeRide's solution can provide reliable perception information even in complex weather conditions.

WeRide has also established an "iron triangle" business model centered on AI technology, OEMs/tier-one suppliers, and travel service platforms: WeRide provides technology and then cooperates with OEMs, tier-one suppliers, and travel platforms to promote the implementation of autonomous driving technology at a low cost and quickly.

WeRide follows the "iron triangle" strategy in all cities, achieving strategic synergies with car companies, suppliers and travel platforms and establishing joint ventures, so that they can always find the most suitable resources. At the same time, because the joint ventures can generate considerable tax revenue for the location, WeRide can always obtain policy support from the local government.

Through this model, WeRide has reached cooperation with Nissan, Baiyun Taxi Group, AutoNavi, and Yutong Group. All of its Robotaxi are regular taxis that can be metered. In addition to Robotaxi, WeRide has also launched Mini Robobus business. As of February this year, WeRide has safely completed a total of 147,128 trips, served more than 60,000 users, and traveled a total of more than 4 million kilometers without any active liability accidents.

Although WeRide's autonomous driving technology has become increasingly mature, the large-scale commercialization of driverless taxis is not something that can be achieved overnight. Going further does not mean pursuing overnight success. Careful calculations are still needed everywhere so that the company can "stand and see the day when autonomous driving is realized."

Many heroes rise up and form alliances, who can "swallow the world"?

Today, the domestic and international autonomous driving markets are full of players. In China, companies such as AutoX, which focuses on handling complex urban traffic scenarios, Horizon Robotics, which focuses on AI chip research and development, Pony.ai, which has collected massive amounts of complex and extreme scene data in China and the United States, and Zhixing Technology, which focuses on pre-installed system solutions in the field of autonomous driving, are developing rapidly. And technology giants such as Baidu and Huawei have already begun to launch their own products.

Overseas, Waymo, backed by Google, just received US$2.5 billion in financing in June. Traditional car companies such as Mercedes-Benz, BMW, and Volkswagen have begun to make efforts. "Outsiders" with strong strength such as Intel and Nvidia have also rushed into the market. There is also Apple, which has been eyeing the automobile industry for many years. According to the habits of this company, once it enters the market, it will inevitably choose full-stack self-research, and will not miss the trend of autonomous driving.

In this emerging market, almost every company has its own strategy. At present, it seems difficult to judge who has the best solution. An industry insider once said: "In the implementation of autonomous driving technology, different autonomous driving companies have their own solutions. These solutions are neither right nor wrong, nor good or bad. As long as they can find the right application scenarios and their own market positioning, they can play their value and be favored by capital."

But all industries have their own rules, and autonomous driving is no exception.

For example, vehicle-road collaboration will be a common development direction, at least Han Xu of WeRide believes so. He said that vehicle-road collaboration is the only way to develop automatic driving. But how to allocate vehicle and road monitoring tasks? On which equipment should detection equipment be installed? How to implement it specifically? I am afraid that different manufacturers will have different ideas.

In addition, regardless of whether there is a competitive relationship between the two companies, "alliances" between the major companies are bound to happen. After failing to cooperate in the development of autonomous driving, Mercedes-Benz and BMW approached Nvidia and Intel respectively to try to make a comeback. On July 17, WeRide also reached a cooperation with Hesai Technology, a well-known domestic lidar manufacturer. The two parties will explore and cooperate in autonomous driving technology, vehicle-road collaboration, business scenarios and other aspects to jointly create a safer and more efficient autonomous driving sensor suite. Similar strong alliances will appear more and more in the future.

As the saying goes, after a long period of division, the world will eventually unite. The current situation of "warring states" may not last long. Mergers and reorganizations within the market may also take place soon. Eventually, the market may stabilize in a situation where three to five companies compete for supremacy. As for whether there will be a hegemon who can "rule the world with long-term strategies", it will take time to verify.

As a winner of Toutiao's Qingyun Plan and Baijiahao's Bai+ Plan, the 2019 Baidu Digital Author of the Year, the Baijiahao's Most Popular Author in the Technology Field, the 2019 Sogou Technology and Culture Author, and the 2021 Baijiahao Quarterly Influential Creator, he has won many awards, including the 2013 Sohu Best Industry Media Person, the 2015 China New Media Entrepreneurship Competition Beijing Third Place, the 2015 Guangmang Experience Award, the 2015 China New Media Entrepreneurship Competition Finals Third Place, and the 2018 Baidu Dynamic Annual Powerful Celebrity.