How AI and HD Maps advance Autonomous Driving
Las Vegas, Jan. 2016 – At CES 2016 visitors had the impression to stand in front of decisive technological disruptions that will determine the future and change society. Artificial Intelligence, the Internet-of-Things, and Autonomous Vehicles were recurring subjects at the show. As already last year the dominant theme at 2016 CES was to create a better world and make all our lives better.
Artificial Intelligence and Cognitive Computing become a key factor in technical progress. Already today millions of data are available in databases and on the internet. Through the Internet-of-Things (IoT) an enormous amount of additional data will be generated. Be it health data through wearables, personal data for adaptation of devices to user preferences, traffic data through V2X communication or through mapping systems for autonomous vehicles. Progress will come with intelligent use of all these data. “Artificial Intelligence and Cognitive Computing will therefore help to make more informed decisions about more consequential matters”, said Ginni Rometty, CEO of IBM, in her Keynote Speech at CES.
With the objective to have in 2020 partially or fully autonomous vehicles on public roads, automotive OEMs presented their next development steps. Key to safely navigate self-driving vehicles are high definition maps and real-time environment scanning capabilities as today’s GPS only provides 10 meters accuracy.
For this purpose real-time vehicle environment data are crowd-sourced with vehicle fleets on the road through real-time scanning with LiDAR sensors and camera systems integrated in (autonomous) vehicles. These data are constantly communicated to the cloud and are compared with high-resolution road map data.
The bigger the fleets are that collect these data the faster accuracy of map data improves.
Therefore, at CES 2016, Mobileye has announced partnerships for their “Road Experience Management” mapping system with GM, VW, and a 3rd partner that was not named, altogether covering 1/3 of the automotive industry. Toyota has announced the development of its own high-precision map generation system. HERE and TomTom have announced the creation of high-definition road maps for highly automated vehicles for which they currently scan roads worldwide.
In addition to map data, vehicles have to recognize objects on roads like traffic signs or other traffic participants and, based on algorithms, decisions have to be taken within milliseconds. In case a vehicle does not know how to react or takes a wrong decision these data are sent to the cloud, treated and deployed to the fleet.
These processes will run on neuronal networks using artificial intelligence to improve driving capabilities of autonomous vehicles over time and make them better “drivers” in low and high speed.
The “CES 2016 Automotive Show Report” edited by StratMa International summarizes these and all other major automotive announcements and technical developments revealed at CES 2016. The report is unique in its presentation as besides illustration through photos on all pages nearly all chapters are supported with video material making the content more comprehensive.