XIAMEN, Fujian – Industry executives and officials who spoke at the 2016 Summit on Auto Data (SOAD 2016) held here on July 21 were bullish on China’s path toward autonomous driving and intelligent transportation but also emphasized that industry stakeholders must work together to move the industry forward and tackle issues and challenges that hinder progress. The event, organized by China Automotive Technology & Research Center (CATARC), had a theme of Auto Industry Transformation in an Era of Big Data. Following are highlights from the event. – Editor
Wu Zhongze: Data integration, quality, standards and security issues need to be resolved
The coordinated innovation of China’s intelligent transportation and auto industry in the era of big data faces three major challenges of data resource integration/utilization, quality/standards and security that must be tackled, according to Wu Zhongze, director general of China Intelligent Transport Systems Association (ITS China) and former vice minister of Science and Technology.
“Traditional automakers are rather closed and data resources are isolated, while transportation industry data sources are complex and scattered,” said Wu. “How the two industries can share and increase comprehensive utilization rate of their respective data resources is a key challenge. The information barrier among industries and companies must be broken and effectively integrated to achieve deep mining and comprehensive utilization.”
Wu disclosed that the Ministry of Science & Technology has allocated a dedicated R&D fund of over ¥3 billion ($461.54 million) to support the development of smart, connected and new energy vehicles. The first batch of 19 projects have recently completed expert appraisal and research work are underway. Wu also indicated that cities including Beijing, Tianjin, Chongqing, Changchun, Wuhan, Nanjing, Wuhu, Changshu, Jiangmen and Baotou are preparing to open pilot projects for intelligent and connected vehicle testing similar to the pilot demonstration zone that recently opened in Shanghai’s Jiading International Automobile City. “This is positive for the commercialization of autonomous driving in China,” Wu told CBU/CAR.
The Ministry of Transport, according to Wu, is working with numerous automakers to draft technology standards related to safety and communications protocol, and ITS China has formed the Highway Vehicle Special Committee and Intelligent Road and Vehicle Coordinated Technology Innovation Industry Alliance to help push the development of intelligence and connectivity in the auto industry.
Wang Jin: Baidu sticking to goal of quantity production of driverless car in 5 years
Wang Jin, senior vice president of Baidu Inc. and president of its Autonomous Driving Division, reiterated the Chinese search giant’s goal of quantity producing a driverless car in five years and commercializing it in three.
“The earlier we achieve this goal, the more lives we save from road accidents which kill 500 people per day in China,” said Wang.
Wang believes a huge divide exists between Level 3 highly autonomous driving and Level 4 fully autonomous driving (based on classification of National Highway Traffic Safety Administration or NHTSA) but with Baidu’s “new intelligence” capabilities, this great divide can be crossed.
According to Wang, artificial intelligence is the core of autonomous driving and Baidu will focus on software capabilities which will account for 60 percent of the value of an autonomous car, which he refers to as software defined vehicle (SDV). It now takes Baidu 1-2 hours, instead of a week (last year) to process over 10+ TB of data per day of autonomous vehicle testing per vehicle, thanks to Baidu’s cloud computing and server processing capabilities. Baidu also now offers what Wang refers to as the world’s top voice recognition capability, which achieves 97 percent accuracy for mandarin speaking in quiet environment and 92 percent accuracy in car. It also offers world leading traffic environment identification accuracy of 89.32 percent (even higher than human recognition accuracy, which is 89 percent) and is the first company in China to offer 3D high precision mapping that is able to pinpoint location accuracy to within 10 cm.
Wang said that the creation of a driverless car ecosystem must involve the integration of stakeholders including national and local governments, automakers, research institutions, mobility service providers, insurance companies and sensor/processer/server manufacturers.
Wei Dong: Big data is about freshness and accuracy
Wei Dong, president of Automotive Business of AutoNavi Information Technology Co., Ltd. and vice president of Alibaba’s Mobile Business Group, believes big data is not about “big” but rather about freshness and accuracy.
“AutoNavi’s goal is to make everyone smarter by providing the most fresh and accurate map data,” said Wei, who touts the Baidu competitor as a “three in one” company that combines mapping data, application and cloud service that offers traffic condition navigation rather than simply traffic route navigation.
One of the reasons that AutoNavi is trusted by users is the fact that 70 percent of the mapping data updates are user generated. One example Wei gave is that the over 40 million packages generated on Alibaba’s Tmall and Taobao e-commerce platforms are all points of interests (POIs) that contribute to mapping accuracy. AutoNavi’s transport information system processes on average 20 billion km of driving data and more than 700,000 road accidents per month. In Beijing for example, it helps save users 680,000 hours in traffic per day.
By the end of 2016, AutoNavi will have collected some 280,000 km of expressway HAD mapping data in China, according to Wei, and the company touts itself as No. 1 in terms of equipment professionalization, high precision mapping data volume and accuracy (within 10 cm).
He Jugang: Chang’an’s No. 1 strategy is intelligent connectivity
Chang’an Automobile is putting intelligent connectivity as its No. 1 strategy, ahead of new energy vehicles, according to He Jugang, vice president of Chang’an Auto Global R&D Center.
“We will adjust our organizational structure based on this strategy and alter our product and service planning accordingly,” said He, who believes that the path toward fully autonomous driving will still be a step-by-step process. “It takes time for the technologies to mature and also consumers to accept them. The key problem to tackle is safety.”
He emphasized however that consumer demand on autonomous driving will not change. Chang’an over the next three years, starting from June this year, plans to develop a prototype V2X vehicle every half-year. By December 2018, the 6th generation of this vehicle is expected to be ready for quantity production.