Research at HKMU is marked by a strong application orientation with a clear goal: to make an impact on society. While it taps into the wealth of the existing faculty, the University is also trying to bring in fresh energy and ideas by recruiting eminent researchers. Among these new recruits is Prof. Roy Vellaisamy who joined the School of Science and Technology (S&T) as Chair Professor of Intelligent Systems in February. A holder of 21 granted patents and more pending applications, he embodies the value of impactful applied research.
With a background in chemical physics and materials engineering, Prof. Vellaisamy is the brain behind numerous practical inventions that find applications in multiple areas. One of his latest projects, for example, contributes to the recovery of industrial waste heat by enhancing the performance of thermoelectric generators (TEGs) — devices that make use of solid-state semi-conducting materials to convert heat directly into electricity. Existing TEGs have two limitations: the constituent materials are brittle by nature, and their power conversion efficiency leaves much to be desired. 'Therefore, we developed a 3D microlattice structure with partially carbonised material for greater mechanical strength and enhanced heat-to-electricity conversion efficiency,' explains Prof. Vellaisamy. The research paper concerned is published in the journal Nature Communications.
Materials engineering is about studying the properties of existing materials with the aim of improving their stability and reliability. It is therefore relevant to all kinds of practical devices. Prof. Vellaisamy's broad interests in intelligent systems include AI-based point-of-care diagnostics, an increasingly valued component in modern healthcare. In S&T, he is working with the machine learning team on data analytics. 'System development is multi-disciplinary and involves various parties,' he elaborates. 'At the molecular level, we engineer sensors, which produce a lot of data. Colleagues with a machine learning background help analyse the data. The results are then fed back to our systems to make them intelligent.'
The materials engineer is no stranger to interdisciplinary research. While his lab facilities have yet to be set up at HKMU, he has already embarked on his first mission to leverage the synergies of multiple Schools to bring direct benefits to society. One of his first projects concerns a diagnostic device for early detection of prostate cancer. Prof. Vellaisamy is involving two Schools in the implementation of the new device. 'To make an immediate impact, I can't work alone,' he expresses. 'I'm working with the School of Nursing and Health Studies to recruit participants to try out the device and collect data, and with colleagues from the Social Sciences Department of the School of Arts and Social Sciences to design a questionnaire to find out what needs to be fine-tuned from the end-user perspective.'
As the leader of the molecular electronics group in S&T, Prof. Vellaisamy is also looking to expand the University's research portfolio and diversify the research student body. 'I'm looking forward to supervising projects on high frequency terahertz materials, and I'm hoping to bring in students from Europe who have a strong foundation in engineering,' says Prof. Vellaisamy, who has spent most of his academic life in Hong Kong and just returned from a three-year teaching stint in Glasgow. He is of the view that cross-cultural academic exchanges will benefit HKMU students and is exploring the possibility of establishing joint postgraduate programmes with overseas universities.
Prof. Vellaisamy is an ethnic Tamil. His hometown and its vicinity are the cradle of many influential engineers, including Google CEO Sundar Pichai. The professor attributes this phenomenon to a strong tradition in learning mathematics. 'Many people try to avoid it, but mathematics is very important,' he says. In his opinion, what Hong Kong needs is not more academic drilling, but an educational approach that boosts students' confidence and soft skills in general. 'While teaching in the UK, I found that students from Hong Kong generally performed very well in the first two years, but they suddenly felt lost in the third year, when more soft skills were called for,' Prof. Vellaisamy recalls based on his observation. Next, he will work with the School's STEM team on strategies conducive to training the next generation of engineers.
Like their fellow graduating students in the Department of Creative Arts, student Kelvin Mak Ka-wai and his teammates put their best foot forward for their final-year project (FYP) several years ago. Entitled 'VR Driving School', the FYP aimed to develop a virtual reality (VR) system that provides simulated driving lessons and mock tests for learner-drivers. When the prototype went on public display at the HKMU Creative Arts Graduation Show 2021, it was given another life. It was first spotted by Lee Kin Driving School, which offered to back a preliminary study on scaling up the system. On this basis, a proposal for a more sophisticated VR-based driving training system was developed, which won another grant from the government's Smart Traffic Fund earlier this year. Thanks to the combination of industry, government and internal funding, the students' prototype is now being expanded into a specific system.
Kelvin's team was in the Bachelor of Arts with Honours in Computing and Interactive Entertainment (CIE) programme, a joint offering of the School of Science and Technology and the School of Arts and Social Sciences (A&SS) that trains technical artists. Dr William Lai, the team's supervisor at A&SS's Department of Creative Arts and the principal investigator of the funded project, had recognised the FYP's potential from the very beginning. Knowing that a complete system would be beyond the scope of the FYP, he had encouraged the team to apply for funding from the then Open InnoChallenge, the University's in-house start-up support scheme. 'We did get the initial proof-of-concept support but didn't make our way to the finals because we weren't that good at sales pitching. We were amazed to learn that Lee Kin Driving School was interested in our project!' says Kelvin, who immediately jumped at the opportunity as soon as he learnt that their FYP was to be developed into a funded project. After a year in gamification, he returned to the University to join the research team as a research assistant.
The initial idea of the FYP was sparked by the team's own knowledge in driving as well as the development of the eye-tracking technology, which was newly introduced to VR headsets. 'We noticed some challenges in learning to drive in Hong Kong, such as the existence of prohibited zones for learners and inadequate driving instructors, so we tried to develop a solution to the problem,' explains Kelvin. Other than providing guidance on the operation of the vehicle, an important role of the driving instructor is to ensure that the learner is keeping a watchful eye. With a built-in eye-tracking script, the VR Driving School is able to detect where the user is looking at and provide feedback signals.
The prototype was constructed based on the Ho Man Tin area in the University's vicinity, where one of Hong Kong's busiest driving test centres is located. The process of reconstructing the road conditions of the driving test route was painstaking. 'Simply building the railings took us two weeks of work day and night!' recalls Kelvin. Developing the traffic rules took them another few months. By the time the project was due, the team had managed to complete a third of the rules, create the first four to five driving lessons and get a few friends to test out the system.
Obviously, a major limitation of the original FYP was scalability, as the 3D models had to be constructed manually. Scaling up the project to cover all driving test routes in Hong Kong warrants an automated process. This is made possible by the extra funding, which is providing for the acquisition of a road scanner for street data collection. Another quintessential improvement over the FYP is the introduction of AI-generated hazard situations that learner-drivers may not encounter in physical lessons. 'We're also purchasing a motion chair to better simulate the driving environment,' adds Dr Lai.
Dr Lai is very pleased with what the student team has accomplished. They have perfectly lived up to the goal of the FYP course. 'We have two major criteria in deciding whether to give the green light to a project proposal — whether there's a reasonable rationale, and whether the project is feasible,' Dr Lai elaborates. 'We won't approve a project just because students find it interesting. We want our students to observe problems and try to solve them by what they've learnt.'
Thus, the FYPs coming from the CIE programme are always practically oriented. According to Dr Lai, it is not uncommon to see a few graduation projects exhibiting good potential every year, but whether any can be further developed and eventually commercialised depends on numerous factors. 'First, the students themselves may have their own plans. Second, obtaining funding is no mean feat,' he expresses. 'But I'm glad the University is now placing an emphasis on knowledge transfer. I hope we'll get even more support in this area.'
