ITRI partners with NVIDIA to develop self-driving and AI technologies

E160922Y5・E160921Y5 Oct. 2016(E203)
On September 21, 2016, ITRI (Industrial Technology Research Institute of Taiwan) signed an MOU with NVIDIA Inc. to form a partnership of artificial intelligence (AI) development and application strategy that integrates the strengths of both sides to develop deep learning technology, AI-based self-driving cars, and smart robots.  This partnership will lay a foundation for using and applying artificial intelligence to different areas.  

According to the general manager & vice president of sales and marketing of NVIDIA Inc. in Taiwan, NVIDIA has created and established a peer-to-peer learning network and also the DriveWorks, a software development kit (SDK) for driverless car developers as the best platform for deep learning.  By combining ITRI’s autonomous development system and vehicle system software, ITRI and NVIDIA work together to both enhance their respective self-driving technology and move the domestic automobile industry to the future of automation.  

Mr. Hu Jwu-Sheng, general director of Mechanical and Mechatronics Systems Research Laboratories of ITRI, indicated that ITRI has been developing self-driving vehicles for more than one year, and ITRI’s specialties include information, communication, electronics, and mechanics.  In addition, ITRI has first started their own development and then worked with National Chiao Tung University and National Tsing Hua University to further form a larger alliance.  This duo plan between ITRI and NVIDIA no doubt serves as an enhancer for the development of the technology industry.  NVIDIA’s deep learning computing chips are used to implement large volume of image computations.  Also, with NVIDIA’s cloud computing server, these images and computation results are all stored on the cloud.  

This strategic partnership is to develop self-driving and AI technologies and also to gather all learning experiences for self-driving cars on the cloud database to facilitate deep learning.  The accumulated technical energy of deep learning and artificial intelligence will apply deep learning technology not only to enhancing the intelligence of industrial robots but also to service robots to create human-computer interaction.  It will also upgrade Taiwan’s industry.  

After self-driving identification testing for months, ITRI’s prototype autonomous car has been able to travel on the road smoothly.  In the early stage, self-driving technology will first be brought to those autonomous vehicles operating on fixed routes at slower speed, such as, shuttle bus at amusement parks, high-speed rail, and track inspection cars for MRT.  It is also expected that the automated shuttle bus service within ITRI be implemented and completed within the coming several years, so that the automated shuttle bus can automatically decide the routes, allow passengers to get on or off at fixed stops, and also return to the nearest stop.  Self-driving technology will be applied to and brought into everyday life.  (September 2016)
/CCS

TIPLO ECARD Fireshot Video TIPLOBrochure_English TIPLO News Channel TIPLO TOUR 7th FIoor TIPLO TOUR 15th FIoor