CS SEMINAR

Design Automation of Cyberphysical Systems: System-Level Approaches for Energy-Aware Electric Vehicle Design and Management

Speaker
Naehyuck Chang
Professor
Korea Advanced Institute of Science and Technology

Chaired by
Dr Tulika MITRA, Provost's Chair Professor, School of Computing
tulika@comp.nus.edu.sg

27 Oct 2017 Friday, 10:30 AM to 12:00 PM

Video Conference Room, COM1-02-13

Abstract:

Why do we drive electric vehicles? It is not easy to say that electric vehicles are higher performance compared with similar price range of internal combustion engine vehicles. There are financial benefits including Government subsidies and tax deduction, which cannot be sustainable. A low maintenance cost is a good advantage, but vehicle depreciation is a big question. Therefore, environmental friendliness should be one clear motivation to drive electric vehicles.

However, electric vehicles are only "zero exhaust emission" because of tire and brake emissions, which occupy a large portion of the total vehicle emissions. Even putting aside the tire and brake emissions, electric vehicles still contribute to significant amount of pollution because of the source of electricity. Electric vehicles produce less than a half of equivalent exhaust emissions compared with gasoline vehicles and not much different from that of hybrid vehicles. Higher MPGe (mile per gallon gasoline equivalent) of electric vehicles can largely mislead the energy efficiency when it comes to "well to wheel" efficiency taking the entire energy ecosystem into account.

It is challenging to make electric vehicles more fuel-efficient because the key powertrain components are already highly efficient, and therefore, there is a very narrow headroom for further enhancement. Consequently, the challenges for extended range of electric vehicles end up with deployment of more lighter materials, which directly impacts on the manufacturing and repair costs, and it may make actual cost of ownership very high. Extended driving range of electric vehicles is one of the most demanding requirements of the current and potential electric vehicle owners, but use of a larger-capacity battery pack makes the vehicle curb weight heavier and thus the fuel efficiency worse.

In this talk, we introduce system-level solutions to enhance electric vehicle fuel efficiency with the current powertrain technologies. First, we develop an instantaneous power consumption modeling of electric vehicles by the curb weights, speed, acceleration, road slope, passenger and cargo weights, motor capacity, and so on, as a battery discharge model. We ensure the model fidelity as we fabricate a lightweight custom electric vehicle perform extensive measurement. The model fidelity enables us to achieve a more accurate range estimation.

We attempt both design and runtime energy optimization using the electric-vehicle-specific energy characteristics. We emphasize that electric vehicles show completely different fuel consumption behaviors from internal combustion engine vehicles due to the significant discrepancy in the drivetrain. We introduce minimum-energy driving methods for electric vehicles, which are largely different from eco-driving methods of internal combustion engine vehicles. We also propose a rapid energy-aware electric vehicle synthesis that allows users to quickly customize their own electric vehicle powertrain specification without understanding the technology.

Finally, we also give a heads up of an application-specific (extreme off-road driving) electric vehicle design challenges. We proved electric vehicles are capable, efficient and promising on extreme off-roads through real trail run tests.

Biodata:

Naehyuck Chang is a Full Professor at the Department of Electrical Engineering, Koreadvanced Institute of Science and Technology (KAIST) from 2014. Before he joined KAIST, he was with the Department of Computer Science and Engineering, Seoul National University from 1997 to 2014. Dr. Chang also served as a Vice Dean of College of Engineering, Seoul National University from 2011 to 2013. His current research interests include low-power embedded systems and Design Automation of Things such as systematic design and optimization of Cyberphysical Systems.

Dr. Chang is an ACM Fellow and an IEEE Fellow for contribution to low-power systems. He was the Chair of the ACM SIGDA (Special Interest Group on Design Automation) and now the Past Chair of ACM SIGDA. Dr. Chang is the Editor-in-Chief of the ACM (Association for Computing Machinery) Transactions on Design Automation of Electronics Systems, and an Associate Editor of IEEE Transactions on Very Large Scale Integration Systems. He also served for IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, IEEE Embedded Systems Letters, ACM Transactions on Embedded Computing Systems, and so on, as an Associate Editor.

Dr. Chang is (was) the General Co-Chair of VLSI-SoC (Very Large Scale Integration) 2015, ICCD (International Conference on Computer Design) 2014 and 2015, ISLPED (International Symposium on Low-Power Electronics and Design) 2011, etc. Dr. Chang is the Technical Program Chair of DAC (Design Automation Conference) 2016. He was the Technical Program (Co-)Chair of ASP-DAC (Asia and South Pacific Design Automation Conference) 2015, ICCD 2014, CODES+ISSS (Hardware Software Codesign and System Synthesis) 2012, ISLPED 2009, etc.

Dr. Chang is the winner of the 2014 ISLPED Best Paper Award, 2011 SAE Vincent Bendix Automotive Electronics Engineering Award, 2011 Sinyang Academic Award, 2009 IEEE SSCS International SoC Design Conference Seoul Chapter Award, and several ISLPED Low-Power Design Contest Awards in 2002, 2003, 2004, 2007, 2012, 2014, and 2017.

Dr. Chang and his colleague introduced the world's first extreme off-road electric Jeep Wrangler and recently finished the Rubicon Trail run for the first time in the history by an electric Jeep Wrangler with just a single recharging at the Rubicon Springs Basecamp.