As early as Apple's investment, people have speculated that Apple is making market reserves for Apple Car, and the "Machine Learning Research Institute" will be able to integrate and technically support Apple's products. Not only is it dropping, Uber also broke news of orders for 100,000 smart cars, and it is easy to get a good ecological entrance for LeTV. In fact, in 2015, China Shenzhou Special Car and Shenzhou Car Rental also made a lot of technical investment in artificial intelligence driving. In this regard , Lei Feng network (search "Lei Feng network" public concern) in the CCF-GAIR after the interview with the China National Taxi Chief Information Officer Liu Yaxi, he disclosed the China AI smart car travel plan.
China Car Rental Wants To Become Cloud AI IntegratorShenzhou is actually a more complex group that covers all aspects of travel. The Group’s earliest establishment was a car rental in China. It is now a car rental fleet in China and even Asia. It covers 74 cities across the country and owns more than 100,000 free vehicles. However, following the trend of the sharing economy, Shenzhou Car Rental began a new field from last year. The Shenzhou Special Vehicle rented 4,500 vehicles from Shenzhou Rent-a-Car to specialize in providing B2C services.
Now, Shenzhou’s special car is driven on the client through application integration, which integrates the car rental, the special car, and the safe driving behavior of the current car into a real parking lot. The all-time AI that can be used, including the auto-run of the car, will become the supplier of smart travel in the future. Liu Yaxi said: They want to build an open smart car operating system so that all those who participate in the system can gain something.
The status of China's autonomous driving technology research and developmentLiu Yaxi said that in terms of cognition, autopilot can increase the efficiency of taxis. In the planning of smart cities, the autopilot-monitored data can be shared with the government and the transportation system, so they started to increase their investment in this area. .
In November 2015, a laboratory was established in Silicon Valley. Currently, there are more than a dozen R&D personnel who specialize in the research and development of smart driving technologies. This system is designed to allow faster integration of existing data and future interventions. Focus on integration, focusing on three areas:
Effective processing of existing car data, data analysis
Developing artificial intelligence and deep learning systems, autopilot is the direction of its application
The establishment of ecological systems, in-depth study of data integration and standardization
Now that artificial intelligence and smart driving are hot spots, Shenzhou has also opened a window of cooperation with globally renowned manufacturers to find operating points and entry points for smart driving in practical applications in China. However, Liu Yachen mentioned that google's driverless cars do not understand the road conditions of China in the United States for 100 years, so the domestic car will have to look at how to make the data more complete and open.
From sharer to shareIn the huge market of autopilot, there are both traditional car companies, IT companies, startup companies, and travel service providers. Car rental in China needs to find a new position in the ecosystem. The current competitive advantage of China is:
Missing operators and integrators in the autopilot system
Lots of car networking data
Have started to do it practically
In the shared economy, cars are shared, but drivers are not. The manned driving economy is a shared economy of people, not something like Airbnb. There is a limit to this. It depends on the driver's time. This is not to reduce the marginal cost but to lengthen the working hours of the driver through the sharing of time. There is a limit to the scale. What is needed in the future is to convert people’s time into machines. People’s time is limited and they should do more valuable things.
The computer can win Li Shishi, and it can also win the tricycleIf you want to see a self-driving car, the first-line car maker can make it on the road within 3 months. However, there are three problems: high prices, insecure data, and inconsistent standards.
China now stipulates that if the car does not achieve AEB (Automatic Brake System) in 2018, it will not achieve a five-star rating. At present, AEB has been installed in Volvo, BMW and other vehicles, so it is said that 2020 should be the time to fully drive the car. In 2030, humans will rely on autonomous driving and become fully commercialized. If you still drive your own car in 2040, it may be like riding a horse on the road.
Liu Yaxi said: “Computers can win Li Shishi, and they can also win tricycles.†Now they hope to share these data and related results with these R&D institutions in the industry and truly have a clear understanding of the Chinese system. They also believe that the entire smart car factory, the entire automatic driving is not a pure behavior, not your own research and development of such a set of OK, it has an end-to-end behavior, sharing can be optimized in order to effectively complete your automatic driving work .
