Fresh off his recent elevation to the status of IEEE Fellow, we sat down with Prof Tony Quek, Associate Head of the ISTD Pillar, to catch up about what he’s been up to and how SUTD is preparing students for the future.
You’ve been recently recognized as one of Clarivate Analytics’ 2017 Highly Cited Researchers in the field of Computer Science. What makes your work so compelling?
Prof Tony Quek (TQ): There are two different approaches which I have been trained in. In Japan, it is very practical and focused on solving industry problems. At MIT, it’s much more technical. They look at the fundamentals, which is much more advanced but might take 20-30 years to solve some of the problems they are working on.
“I try to put the two together – researching important problems that can advance the field, while also having practical applications that the industry is interested in. This is also something I try to impart to my students and researchers as well.”
An example is my research on heterogeneous & wireless networks.
How would you explain this research to a layperson?
TQ: Essentially, wireless networks are increasingly heterogeneous which means devices run on more varied WIFI and cellular standards. As watches, autonomous vehicles, smart cameras and IOT devices come online, each needing more and faster data, we will need a better infrastructure to support everything.
What I’m trying to understand in my research is to figure out how to design this network infrastructure. It used to be done by trial and error. For each network, we would deploy, take measurements and then make corrections. The problem is that these new devices and services have different networks standards and requirements, making it exponentially more complicated and expensive. So how can we create a framework to figure out the best configuration, we can deploy, so that as a user your devices just work, no matter which standards they require? That’s what I’m trying to figure out.
The other problem with the trial-and-error approach is that say you deploy and optimize a network for Singapore. Would it work in Shanghai? What about Jakarta? If we can create a framework to guide this process, we can potentially save a lot of time and resources.
What does this mean for the real world?
TQ: There is a lot of industry interest in this area from companies like Qualcomm, Huawei, LG and Samsung. In this new technological generation, they need to decide which new technology to adopt, and how much capital investment to make.
Because there would be so many combinations, it would be prohibitively expensive to take the trial-and-error approach for each configuration. With a framework, they will be able to optimize the network to provide sufficient coverage and bandwidth for everyone. Imagine Singtel being able to provide twice the data for half the price.
What else are you excited about in this space?
TQ: One of the other things I’m very interested in is Network Intelligence. It’s like an AI in the network. With more and more data being captured by IOT sensors everywhere, you might see some amazing things once some intelligence is applied at a network level.
I’m currently starting work on an Intelligent CCTV system. The problem with current CCTVs is that it’s only after you realize something has happened that you go and review the footage. Sometimes you then realize that the system is faulty, or at best a crime has already been committed. With a smarter system, perhaps you might get a notification, that there is suspicious activity you should be monitoring.
Another advantage of intelligence networks is that we might be able to link up different systems onto the network. Saying you need a new system is easy, but we want to be able to integrate different systems into the same network. We can then plug in cheap off-the-shelf cameras and have them work together with specialized high-end ones as part of an entire network.
This is currently still at an early phase in the research, but given the potential industry and government applications you can see why there is a lot of interest in this.
You mentioned AI – what do you feel about the prospects for SUTD students considering automation and AI?
TQ: SUTD is taking the right approach. I’ve been fortunate to have experienced being trained and working the US, Japan and China, developing an international perspective. This perspective is a critical skill for students to have in the coming years as there are a lot of opportunities that will open in Asia.
“At SUTD, we are making great strides in encouraging a deep international experience through various initiatives, internships and exchange programs.”
Last year, our team kickstarted a program where we had various groups of students make several visits to Alibaba’s headquarters in Hangzhou to learn about their various processes and how they operate.
This was a valuable experience for them, as it not only gave them an international experience, but also exposed them to the entrepreneurial spirit and speed at which technology and business operates at in Asia.
Another such collaboration is our partnership with Zhejiang University. In addition to the five electives on Chinese culture, technology and entrepreneurship my colleagues from ZJU developed for SUTD, I also help to run the SUTD-ZJU Innovation, Design and Entrepreneurship Alliance (IDEA). It is a resource for students, or even faculty and researchers to develop their ideas into something for the much larger Chinese market. I hope that eventually some of these can even spin off into full companies.
“This start-up mindset is something I try to inculcate in my students and my researchers. This is SUTD’s philosophy as well.”
For students, our curriculum trains them to be interdisciplinary, which is what the market needs. They graduate with the mindset that they can work across disciplines, whether it’s with an engineer, architect or mathematician.
What about graduate students or researchers you supervise in the Wireless Networks and Decision Systems Group?
TQ: I take more of a tailored approach to their capabilities and career plan. The group already selects for independent and passionate researchers, so I just fine tune the problems they tackle to their interest and capabilities.
Some people are better at theoretical mathematics and are better at working on problems that might take years to solve, while others are more excited about the implementation and putting theories to the test in the real world. Both are important so we fine tune the experiences to fit their personalities. If they are looking to start up a company after this phase, we might look for something that would give them the relevant experiences and networks.
The thing that doesn’t change is the expectation of high quality work. Just because we’re from a new school, it doesn’t mean that the work we produce shouldn’t be able to compete with MIT or Stanford. We need to keep pushing ourselves to tackle problems that matter.
It is an exciting time and I believe that the SUTD experience helps to build a competitive advantage of being able to work with people from different cultures with an entrepreneurial mindset on top of their technical expertise. They will be well placed to take advantage of the opportunities that come.