MEMBER SPOTLIGHT
RATNAK SOK
This ASME Forward Ambassador is improving ICE and EV sustainability and viability.
Written by Cathy Cecere
ELECTRIC VEHICLES (EVS) ARE GREAT, but “they’re not the only solution to get us to carbon neutrality,” declared Ratnak Sok, associate professor at the Research Organization for Next Generation Vehicles at Waseda University in Tokyo. Sok’s expertise, encompassing electrified vehicles, thermal engineering, CFD, and the internal combustion engine (ICE), has led him to believe in “multi-pathway approach” to meeting the challenge of climate change.
Sok is part of an elite group of global researchers who help promote the ASME’s work with ICE—he is an ICE Forward Ambassador. The carbon neutrality targets set by the countries and governments of the world are critical, ambitious, and difficult, Sok said, concluding, “We cannot rely only on battery electric vehicles. Currently, about 95 percent of the vehicles in the world run on internal combustion engines. So why not continue to develop the engine to make it better?”
EFFICIENCY IS KEY
The world is looking toward e-fuels, ammonia, or hydrogen, to gain a better carbon dioxide advantage faster, he explained. And you can reach your goals “without changing much when it comes to engine, hardware, or current infrastructure,” Sok concluded. This is mainly because when you consider the life cycle of batteries, for example, “raw material refinery and battery manufacturing still use energy from fossil fuel.”
There are countries where EVs are a very good match, he explained. “For countries like Norway or Iceland, it's perfect. EVs make sense in those countries because in those nations’ energy comes 98 percent from renewables.” There are challenges including “enough charging stations” to serve these automobiles and how expensive these EVs are to the ordinary consumer. So costly, in fact, that “most consumers cannot afford them,” he explained.
And finally, passenger cars are not the only means of transportation that need to be transformed. “You have to electrify other transportation sectors,” he explained. This includes heavy duty trucks, off-road, marine, railway, agricultural, and other machinery. “These segments are hard to electrify,” Sok explained.
VIRTUAL SENSORS
Regardless of the engine type, whether an ICE or an EV, the need to be more efficient is key. Physical sensors are commonly used to record performance data of ICEs, for example for online feedback control and calibration. But they’re prone to diagnostic and increased development costs. Lookup tables are commonly used in conventional calibration and feedback control; however, the table parameters increase with the advancement of ICE technologies under transient operations. Consequently, the calibration and control systems are time-consuming.
Ratnak Sok (front right) joins 2024 Waseda University graduates and lead researcher Jin Kusaka (front center). Photo: Waseda Mobility Group
ASME ICE Forward Ambassador Ratnak Sok has developed a virtual diesel engine and virtual sensors to provide an onboard feedback control system that predicts the combustion, performance, and emission of internal combustion engine using neural networks and image processing.
That is where the use of virtual sensors comes into play. One big advantage of digital sensors over the physical kind is durability. “When you put a physical sensor in an engine, you run the vehicle for several years. Those sensors age. And overtime they get damaged by extreme conditions such as, for example, very high speed or high load driving,” he explained.
Sok’s work includes developing novel virtual sensors to address these issues by predicting the combustion, performance, and emission of ICEs using neural networks and image processing/translation. The novel sensors are targeted for onboard feedback control systems under transient driving.
To make the scheme work Sok and his team needed to first devise a virtual diesel engine (VDE) and calibrate it against experimental data taken from a production 2.2 L turbocharged diesel engine. The VDE was used to generate teaching data such as indicated torque, fuel consumption, engine brake thermal efficiency, maximum pressure rise rate, NOx, and CO2 emissions. Next, Sok’s team developed virtual sensors using five machine learning regressors.
“When you use digital sensors with AI you are able to make advancements much quicker,” Sok concluded. The downside of virtual sensors are, however, that AI needs a lot of data. And further, original equipment manufacturer (OEM) data needed to build an accurate model is very difficult to obtain because of confidentiality
Sok plans to extend the scope of his research to compare the predicted and desired rate of heat release using the feedback system. This feedback system will then be used to control the injection pattern, which could yield optimal fuel economy and better emissions of any type of advanced ICEs (fueled by either gasoline, diesel, e-fuel, or hydrogen). According to Sok, this will pave the way for a fast feedback control system and reduced engine calibration and control costs. The same virtual sensing can also be applied to predict dynamic battery responses too. He added that he plans to apply artificial intelligence or machine learning applications to predict EV performance in his upcoming collaborative projects with OEMs, too.
Cathy Cecere is the membership content program manager.
© 2025 The American Society of Mechanical Engineers. All rights reserved.