Driven by scientific pursuits

Jul 28, 2020

(From left) Professor Huang Mingxin and Dr Luo Ping are two of the many young innovators at the HKU Faculty of Engineering.

Their time spent abroad shaped their scientific research career but for two professors from The University of Hong Kong’s Faculty of Engineering, it is their commitment to fundamental research that led to their notable breakthroughs.

Professor Huang’s Super Steel brought him the Global Innovation Award at TechConnect World Innovation Conference & Expo in 2018.

Professor Huang Mingxin from the Department of Mechanical Engineering had spent four years doing both his doctoral studies and research backed by ArcelorMittal in Maizieres-les-Metz, France, the world's leading steel and mining company, in the 2000’s. Originally from the mainland, he joined HKU in 2010 and has since won a string of awards and patents for his inventions in the area of metallic science.

Ground-breaking steel

His latest discovery – Super Steel (also called D&P Steel) – was published in the prestigious journal Science in 2017 and 2020, and brought him the Global Innovation Award at TechConnect World Innovation Conference & Expo in 2018.

"We attained an unprecedented strength-toughness combination which can address a major challenge in safety-critical industrial applications - to attain an ultra-high fracture toughness so as to prevent catastrophic premature fracture of structural materials,” said Professor Huang. “The breakthrough also changes the conventional view that attaining high strength will be at the expense of deteriorating toughness, which invariably leads to the embrittlement of structural materials and greatly limits their application."

The Super Steel has great potential for mass production since it is exceptionally inexpensive and can be developed using conventional industrial processing routes, including warm rolling, cold rolling and annealing, in contrary to the complex process of developing other metallic materials.

Steel production is indispensable to modern living anywhere in the world, and Professor Huang has had long term collaborations with companies in Europe and China, such as BaoSteel, Angang Steel, besides the French company he spent years working for. “Researchers are constantly trying to make better steel to make our lives better,” he said. “Much investment has been poured into research on steel globally but not in Hong Kong as it is a small place.”

But having built up his own lab and research team at the department, thanks to government funding and the support of other funding agencies, he has carved out a distinguished career discovering solutions to issues long associated with the production of steel and other key materials.

“Usually when you make steel very strong, they become brittle, like ceramic. It cannot be used to make cars or planes. My Super Steel is very strong but not brittle, it’s even more ductile than other steel and can avoid the problem of sudden fracture.”

New venture ahead

Professor Huang will lead a team to develop new nanostructural high entropy alloys that have the potential for application in extremely low temperatures or in nuclear power plants.

In January this year, he was awarded an RMB11.07 million funding in the second round of the National Key Research & Development Programmes (NKPs) of the PRC’s Ministry of Science and Technology (MOST) that is open for researchers in Hong Kong and Macau.

The National Key Research & Development Programmes support research and development in areas of social welfare and people’s livelihood, such as agriculture, energy and resources, environment, and health, focusing particularly on key and strategic technologies.

Professor Huang will lead a team of researchers from HKU, the Chinese Academy of Sciences, Dalian University of Technology and Xi’an Jiaotong University with expertise in areas such as metallurgy, mechanics, nuclear technology and ab-initio calculation to develop new nanostructural high entropy alloys that have the potential for application in extremely low temperatures or in nuclear power plants.

The need for a new type of alloy became prominent after the collapse of the Fukushima Daiichi nuclear power plant in Japan in 2011. “The alloy in the shell of the plant had been used for a long time, its risk tolerance was very low and collapsed after a long period of exposure to radiation. It was not ductile enough to resist fracture when the accident happened.”

His team has looked into other useful materials including Nickle free stainless steel, which has applicational value for the watch industry for customers with Nickle allergy; anti-virus steel that can be used for the benefit of public health, for example in toilet flush or lift buttons.

Professor Huang attributes his commitment to research to a free, playful childhood. “When I was small, I just played all the time; my dream was nothing but to be able to play everyday. Maybe it’s in my nature to find something new and interesting every time. Like in research, I will continue with it as long as I find it interesting. I also want to create something new out of it.”

Being awarded fills him with happiness and pride. He is equally thankful for the opportunity of working in his own lab. “I am thankful to our department. Like the entire university, they have changed from being teaching oriented 20, 30 years ago to being research intensive.”

Young Innovator in the midst of perseverance

Dr Luo Ping is one of the 20 Innovators Under 35 for the Asia Pacific Region, selected by MIT Technology Review.

Another innovator from the Faculty of Engineering, Luo Ping, an Assistant Professor at the Department of Computer Science, ranks among the 20 Innovators Under 35 selected for the Asia Pacific Region by MIT Technology Review, a prestigious global news media on important new technologies under the Massachusetts Institute of Technology. Among the past recipients of the honour are Facebook’s Mark Zuckerberg, Google’s Sergey Brin and Larry Page.

Originally hoping to study medicine, Dr Luo embarked on the now immensely popular Artificial Intelligence (AI) research after he visited an AI research institute established by a professor from the University of California, Los Angeles, while still a third-year software engineering student at Sun Yat Sen University, Guangzhou.

During his visiting he was exposed to efforts by scientists there to solve computer vision tasks, such as recognizing a noisy facial image. That inspired him to do computer research. In 2011, he began his doctoral studies at the Chinese University of Hong Kong through the Hong Kong PhD Fellowship Scheme. There he met his supervisor who later founded Sensetime, a company specializing in innovative AI technologies, and recruited Luo as research director.

“We delivered intelligent vision technologies for several industries such as smart cities, smartphones, and autonomous vehicles,” said Dr Luo. His own research focuses on developing computer vision and AI technologies to understand human behaviors such as faces, emotions, actions and social relationships, in order to advance human/AI paired systems that outperform their singular counterparts.

His breakthroughs cover autonomous machine learning, normalisation, and optimisation of deep neural networks. He also contributes many popular benchmarks to facilitate AI and computer vision researches for both academia and industry such as CelebA, the most widely used image database for face image generation, and DeepFashion, the largest image database for understanding fashion trend by analysing clothing images. His input helped companies in delivering online sales, mobile shopping/entertainment, and virtual try-on, as offered on the e-commerce site Taobao.

The core technology of AI


Dr  Luo contributes many popular benchmarks to facilitate AI and computer vision researches for both academia and industry. DeepFashion, the largest image database for understnaidng trend by analysing clothing images, is one good example.

The foundation of the above researches and applications lies in understanding how deep learning really works in computer vision. He recalled with pride that he was the first student in his research group and even in Asia to do research in the area while at CUHK. “After our experiment, students were split into groups by the professor to do deep learning.”

In 2014, his research team broke world record of face identity analysis by surpassing recognition rate of human vision for the first time worldwide in the Labeled Face in the Wild (LFW) Challenge, outperforming Facebook, Google and many other institutions. His specialized area of convolutional neural network (CNN) has important applications in robots and smartphones.

While facial recognition system is increasingly used in different countries, at shops, subway stations, airports and other venues to bolster efficiency and convenience, Dr Luo expects the technology to be widely adopted in about three years, and humanlike robots to become reality in about a decade. 

“In some mainland cities such as Harbin, under an automatic payment system at the metro station passengers do not need to show any card or credit card to enter the train. Because of the facial recognition system, they don’t have to worry about how to pay. Machines nowadays can understand human identity.”

The human brain, he said, which is capable of performing multiple tasks, can be considered as a single general machine learning model. But there is no need to mimic the biological mechanisms of human brains. “With the development in deep learning, we are able to design a mixture of different models with each solving specific tasks. For some specific tasks, machines can do better than the human brain,” he added. “For examples, AlphaGo, a computer program that plays the board game Go, beat top human professional players in 2016[1]; In 2019, another program AlphaStar beat top human players at strategy game StarCraft II[2] and a poker-playing bot Pluribus beat human in six-player no-limit Texas hold'em poker game[3]. And in reality, more than 60% of stock trading in the stock exchange are made by AI software.”

With an interest in teaching and hoping to educate the new generation of AI experts, Dr Luo returned to the academia, joining HKU in 2019. Although there are far more AI companies offering bright career prospects and lucrative pay in China, Dr Luo prefers to remain here.

“The opportunities in Hong Kong for AI are a bit limited but I still want to motivate young people to join the field. There are many good AI companies in Hong Kong and there can be more in the future.”