您现在的位置: 易图通正文

Connected Vehicle and Autonomous Driving | EMG User Conference 2018

发布时间:2018-05-02  阅读次数:565    
On April 26, 2018, EMG User Conference 2018, was held successfully in China World Summit Wing in Beijing. Over 400 people from near 50 auto makers, most major system vendors, many famous AI companies and other partners within the ecosystem joined the conference to share their insights on the hot topics related to connected vehicle and autonomous driving.




EMG CEO Darwin Lu

Mr. Lu summarizes EMG’s achievements in 2017—establishing highly automated HD map production line, providing ADAS verification & simulation service for AV, delivering more comprehensive connected car service solutions, launching online world map to serve global users and ranking 3rd in OEM map vendors with 26% of OEM in-vehicle navigation map market share.

He points out that EMG is a neutral mapping company and will focus on serving its customers without intervention from its shareholders. With a philosophy of “cooperation, openness and sharing”, EMG upholds an open mind to deal with the issues of cooperation and competition. It is willing to cooperate with other companies, even though some of them may sometime become EMG’s competitors.

Mr. Lu believes a new business model of connected vehicle service is evolving and it will create more values for auto-makers and value chain providers. He reveals that EMG is also focusing on R&D of HD map to prepare for the upcoming opportunities of L2.5-L3 autonomous driving.



EMG Vice President Vincent Yang

Mr. Yang shares EMG’s roadmap of autonomous driving. It plans to combine the strengths of two technical approaches, one that relies on heavy sensors including LiDar, Camera, IMU, RTK GPS etc. and the other that only relies on light sensors like camera and IMU, to meet the requirements of quick covering, real-time update and positioning accuracy of HD map.

He also addresses that HD map is not only useful for autonomous driving but also can be widely applied in intelligent transportation and other fields.




EMG Vice President Dr. Hanke Zhao

Dr. Zhao says, with tens of millions of vehicle probe data and hundreds of millions of mobile probe data, EMG is building a “live road network” with three layers, base road network layer, dynamic layer and real-time layer. In base road network layer, EMG’s map has covered near 7-million-kilometer roads and will cover 8.5 million kilometers by the end of 2018. Dynamic layer includes information of road networks, such as regional traffic restriction, reversible lanes and road constructions etc. By July 2018, regional traffic restriction information will cover all cities with documented traffic restriction rules. In real-time traffic layer, over 120 cities, all highways and 85% of national & provincial roads have been covered.

Dr. Zhao also stresses that with its partners, EMG has already created a competitive connected car service solution which has been chosen by several OEMs.



Tsinghua University Director of Center for Intelligent & Connected Vehicle and Transportation Prof. Keqiang Li

Prof. Li says, the industrialization of intelligent & connected vehicles urgently needs to build up five base platforms, including HD Map Data Base Platform, Cloud Control Base Platform for Intelligent & Connected Vehicles, Cloud Control Base Platform, in-Vehicle Intelligent Computing Base Platform and New Intelligent Vehicle Terminal Base Platform. The industrialization of these five base platforms will bring synergy to serve for the national strategy of intelligent & connected vehicles.




Ygomi Chairman (founder of NAVTEQ) T. Russell Shields

Mr. Shields thinks, precise localization is a key to highly automated driving. Absolute position is not needed, but precise, meter-by-meter, lateral and longitudinal localization to 10cm within the lane is necessary.

Mr. Shields points out that the road database for highly automated driving is not a map, because it has no street names and addresses, no POIs and is only used by machines. He emphasizes that as the road database for highly automated driving must be completely reliable, it must be created only by software without any human intervention.



Mapbox Vice President Alex Barth

Mr. Barth reveals that each day, Mapbox collects 225 million miles of probe data and use them to update a live map in real time. He demonstrates to the audiences how Mapbox applies its new mapping technologies to unlimited usage cases.


Altran UK Automotive & Product Unit Chief Engineer Michael Welsh

Mr. Welsh addresses the complexity of connected vehicle and autonomous driving and emphasizes the software framework in the vehicle needs to be changed. Altran has developed coherenSE@ platform with new SOA frameworks. It can avoid functional domain names to help manufacturers and suppliers get rid of the limits of software and hardware management within the production chain of intelligent vehicles.

 

Momenta Co-Founder and CTO Dr. Gang Sun

Dr. Sun says, autonomous driving will largely improve the travel safety and lower the travel costs. In order to commercialize autonomous driving as soon as possible, Momenta adopts a progressive approach, at current stagethat is visual recognition technical path at a relative low cost. This path requires a strong capability of AI algorithm which is exactly Momenta’s core competence.



GNSS and LBS Association of China Chief Scientist Chong Cao

Mr. Cao says, looking back to the history of US loosening SA and China relaxing the restrictions of digital navigable map, it can be speculated that precision is a false proposition and it will not be an obstacle of the development of autonomous driving. People should focus on reliability, safety and system of autonomous driving and tackle the difficulties of fusion of technologies, policies, regulations, standards, verifications, costs and safety.

Mr. Cao suggests that the undergoing formulation of Satellite Navigation Regulation will be a very good chance for introducing autonomous driving to related regulations. He points out that in order to promote government legislation, the industry should provide the government solid and pertinent research reports to help it make right decisions. Industrial association like GNSS and LBS Association can play an important role.


Z Ventures Group Dr. Vincent Tao

Dr. Tao believes that the future of the mapping industry belongs to AI and spatial big data. Mapping is transforming from for Earth to for human and to for machine. Correspondinglythe source of needed data is transforming from mapping and surveying, to big data of human behaviors and to big data of robot behaviors.

 

After the speeches, six guests from a map provider, an auto maker, two AI companies, an Internet company and a university had a panel discussion on “Opportunities and Challenges of Connected Vehicle and Autonomous Driving”.

Do you want to know what they have said in the panel discussion? Coming soon!