Plenary Lecture

Disruptive Technologies for Electric Vehicle: Neural-network based AI vision and Wireless Power Transfer

Professor Il Song Han
Graduate School for Green Transportation
Korea Advanced Institute of Science and Technology
Daejeon, Korea
E-mail: i.s.han@kaist.ac.kr

Abstract: Cars and commercial vehicles are edging the way towards electric due to the environmental benefits and cost savings. While Electric Vehicle’s quiet motor is a major environmental attraction, it also creates a dramatic hazard for all road-users particularly the elderly, the young, cyclists and other vulnerable road users. The advanced safety feature for vulnerable road-users is a critical technology in demand, as the up-to-date technologies of driverless car and electric vehicle are still restricted by those accidents. Wireless power transfer is another technology to shift the paradigm in Electric Vehicle, which offers the autonomous, safe, efficient and convenient option to plug-in charging.
The physiological studies about visual cortex have inspired a new way of implementing the robust Artificial Intelligence of vision by Neuromorphic neural network mimicking brain functions. We studied the Neuromorphic convolutional neural network based on Hubel & Wiesel’s experimentation and Hodgkin- Huxley formalism. The robust performance of visual detection was applied to detect the vulnerable roadusers on the road. The performance and engineering of Neuromorphic processing were tested and demonstrated by the detection of pedestrians and cyclists for vehicles on the road. The real-time operation engineering prototypes were evaluated for the feasibility of various applications, based on embedded system of FPGA and GPU.
The wireless power transfer has been investigated with the new material for its feasibility to Electric Vehicle, while the wireless powered Electric Bus was successfully evaluated for the public transportation. The new magnetic material was evaluated for its advantages in Electric Vehicle, for its further robustness in mechanical stress and electrical capacity. The proposed nano-structure and fabrication process demonstrated the feasibility of wireless power transfer for Electric Vehicle.
Our conclusion is that the Neuromorphic neural network of AI vision and the robust magnetic material of wireless power transfer to become two seminal technologies for Electric Vehicle.

Brief Biography of the Speaker: Prof. Il Song Han is an invited professor at the Graduate School for Green Transportation at Korea Advanced Institute of Science and Technology. His research areas are analogue-mixed VLSI design, neural networks, wireless power transfer of electric vehicle, neuromorphic artificial intelligence for automotive safety and driverless car, focusing on technology innovation by neural networks and ASIC. His most recent work includes development of new engineering prototype based on neuromorphic vision for the automotive industry. He has the diverse experience in both the academia and business, having worked in telecommunication and ICT sectors over 30 years in both UK and Korea. He is the co-founder of technology start-up in UK, for promoting the artificial intelligence in safety technology.

Bulletin Board

Currently:

The Conference Guide is online.

The organizing committee is accepting papers and special session proposals.


Plenary Speakers

WSEAS Main Site

Publication Ethics and Malpractice Statement