Kai-Fu Lee, an artificial intelligence scientist and chairman and CEO of Innovation Works, once predicted that the real realization of unmanned driving will take 15-20 years. It can be seen that there are already cases of unmanned driving on a small scale. At a time when demand is growing, the transformation of autonomous driving in three to five years is worth looking forward to.
Recently, the investment boom in autonomous driving has continued. According to reports, SoftBank may finalize a new investment, leading the investment of 300 million US dollars (about 2.1 billion yuan) in Didi’s autonomous driving subsidiary. If the transaction is concluded, this will be SoftBank’s fourth financing to Didi. In addition to reflecting Softbank’s contribution to Didi’s snow, it also reflects the current capital’s recognition of the future potential of autonomous driving.
Investment boom coming?
The commercialization of autonomous driving has always been the focus of attention, and the sudden epidemic has instead made the outside world see the need for unmanned services, such as unmanned taxis, unmanned logistics, unmanned distribution, and so on. At the same time, it also promotes the research and development of technology by autonomous driving companies, so as to achieve the best results of safety, efficiency and low cost. In addition, the release of new domestic AI infrastructure and autonomous driving standards has also injected a booster into the industry.
“I think the development technology of autonomous driving has reached a new stage, and basically the investment circle will also believe that autonomous driving will definitely be put into operation in the next three to five years and will be widely used in a wide range of applications.” Han Xu, founder and CEO of WeRide In an interview with the 21st Century Business Herald reporter, he said: “This is no longer an investment full of high risks in the past, but everyone has seen this track that requires capital very much, and what is needed is to expand operations and technical stability. I predict there will be a lot of integration in autonomous driving in 2020.”
Han Xu also revealed that the company was originally going to conduct a B round of financing, but due to the impact of the epidemic, the pace of the B round of financing may slow down slightly, but recently, there have been continuous financing news, and the capital market has become hot again.
Wu Gansha, CEO of Uisee Technology, told the 21st Century Business Herald reporter: “The financing started well at the beginning of the year. The main reason is that I think everyone has stronger confidence in the real commercialization of intelligent driving. Generally speaking, a high-tech It takes about five or six years for the technology to come out from the laboratory to the real large-scale commercial application. We judge that 2020 should be the first year for the real large-scale commercialization of unmanned driving.”
At present, the epidemic has catalyzed the further implementation of autonomous driving. For example, some companies have used autonomous driving to transport materials such as masks, disinfectants, and catering in epidemic areas. There are also frequent cases of unmanned vehicle logistics delivery in airports and factories. Unmanned vehicles have their advantages in reducing the workload of medical staff and ensuring safety and convenience.
Uisee Technology, which focuses on unmanned vehicle logistics, has recently received demand from customers to increase orders. “Orders are increasing at an order of magnitude, and we also see that there may be some retaliatory rebounds in demand in the second half of the year.” Wu Gansha said. At present, UISEE has normalized operations, one is the airport logistics at Hong Kong Airport, and the other is the factory logistics done by UISEE and SAIC-GM-Wuling in its Baojun production base.
The key is to reduce costs
In Wu Gansha’s view, to cut in unmanned, we need to look at three factors. The first is whether it is high-frequency rigid demand; the second is that the scale is feasible and replicable, and the security officer can be removed; the third condition is The economics are better, customers can make money, and self-driving companies can make profits. He said that by the end of next year, he expects the company to achieve normalized operations on a scale of 1,000 units and achieve 200 million revenue at the same time, so as to truly realize commercialization in a financial sense.
And WeRide Zhixing opened its self-driving passenger operation in December 2019. At present, users can enjoy taxi services on the 144-square-kilometer urban open roads in Huangpu District of Guangzhou and Guangzhou Development Zone. However, limited by national laws, a security officer is still required to ensure complete security. Due to the outbreak of the epidemic, WeRide began to resume Robotaxi’s passenger-carrying operations in Huangpu District, Guangzhou on March 2.
Han Xu said that if it is really possible to achieve pure unmanned driving, it is very suitable for transporting suspected patients, because there is no other driver in the self-driving unmanned vehicle except the patient, which can also avoid cross-infection of the next patient. However, it will take two to three years for the popularization of pure unmanned autonomous driving, and WeRide will also accelerate the research and development of technology to achieve better stability.
He told reporters: “We decided to do our best to continue to expand operations in Guangzhou and deploy hundreds of Robotaxi in 2020. In 2021, we look forward to pushing the policy to remove security guards in the areas we are familiar with, and in 2023, we hope to have most of Guangzhou in Guangzhou. area, and implement Robotaxi’s unmanned passenger operations.”
Some people in the industry also pointed out to reporters that cost reduction is the key to commercialization. At present, regardless of algorithms and computing power, as well as external infrastructure, there is a trend of cost reduction. Wu Gansha also mentioned that UISEE has rich scenarios, and the migration cost between different scenarios can be very low. “In terms of technology, we actually have a good fit with specific models, different technologies and business scenarios. In terms of algorithms, we also try our best to generalize through machine learning methods.”
Reducing costs is the key to the commercialization of autonomous driving. At present, regardless of algorithms, computing power, and external infrastructure, there is a trend of cost reduction… However, it should be noted that there are still many manufacturers who want to enter the large-scale commercial stage. The hurdles need to be overcome, including technology, industry chain maturity, policy, and key cost issues.
The Links: SKMD150F12 LP121X04-B2