People Behind the Science - Dr Jordan Ivanchev

25.09.2020

 

At TUMCREATE, as we prepare to conclude our research in Phase Two – Towards the Ultimate Public Transport System, we would like to highlight our researchers who continue to make significant contributions to the programme. Through our interview series, People behind the Science, we talk to our researchers to discover more about them and their stories.

 

Dr Jordan Ivanchev, from the team Area-Interlinking Design Analysis (AIDA), joins us in our interview.

 

Jordan grew up in Bulgaria, a small country in Eastern Europe, where he described as a place filled with rich history, beautiful traditions and warm, kind-hearted people. Besides being active in various sports, Jordan is an airplane enthusiast who gets really excited when it comes to flying, or anything that can fly, be it an airplane or a spaceship. In his free time, he would unwind with a group of friends or sits in front of a piano, playing his favourite tunes. 

 

Q: Tell us about yourself and what you do…

I am currently a senior research fellow heading the modelling efforts for our traffic simulator, CityMoS. I have also been part of the Cooling Singapore project, since its beginning in 2016, where I model mobility-related aspects of human behaviour.

 

I majored in Electrical Engineering and Computer Science for my bachelor’s degree in the Jacobs University in Bremen. After that, I received a master’s degree in Spacecraft Technology and Satellite Communications from the University College London (UCL) and a PhD degree from the Technical University in Munich (TUM).

 

Q: What made you come to Singapore, and how has it been like living here?

I like to say that I came to Singapore because of a lucky misunderstanding. The job advertisement that I applied for only mentioned TUM and I assumed the position was in Munich. At the end of my interview, my interviewer said, “Let me tell you a little bit about Singapore”. I got a little confused and asked why that would be necessary and he simply told me that it was because I would have to live in Singapore. That was when I realized I had to move to Singapore for the job! 

 

Despite my initial shock, I didn’t really hesitate much. I gladly took up the challenge, ready for an adventure, when I was offered the position. I remember vividly that he described Singapore to me as the “Asia for beginners”, which I later found to be quite an accurate description. Living in Singapore has massively enriched my perception of the world. Here, I have met and interacted with people from different cultures, made new friends, worked with great professionals and started a family of my own.

 

Q: Who or what inspired you to be in your field of research?

I have been blessed with a few great teachers along my education path, as early as middle school, that have shaped the way I think, reason and approach and define problems. These individuals, together with my father, have taught me that math is not only a universal tool that can be used to tackle challenges from almost any domain but also a universal language that has the power to seamlessly connect on a very deep conceptual level with people from anywhere in the world. I find it quite romantic that something so abstract, rigorous, and formal can be used to create elegant and intricate solutions for complex problems that reemerge from different academic fields and aspects of life. 

 

I started off my research career in the field of machine learning, which is a small subset of the vast field of mathematical modelling. Although he has quite a peculiar personality, I have to say that Juergen Schmidhuber is one of the most inspirational contemporary thinkers in the field and in artificial intelligence in general. I believe that many other machine learning practitioners of my generation were also attracted to the field by his thoughts, ideas, and his unique way of presenting them.

 

Q: What are some of the challenges for you in your field of research?

As a modeller working at the boundary between product and stakeholders, the biggest challenge has always been to convince the users of the value that comes with the application of our models. I believe that defining model performance metrics is a crucial part of research nowadays and that it is best tackled when both researchers and users work on it together.

 

Being able to synthesize all information defining the performance of a model into a single number is truly an art that requires a lot of experience and in-depth understanding of both the research field and real world context of the problem, be it financial, social, political, or environmental.

 

Q: Tell us about your work and journey in TUMCREATE…

I joined TUMCREATE in 2013 and am probably one of the people who have been with the company for the longest time! After finishing my doctorate thesis in 2016, I assumed a more versatile role that included being involved in other work on top of my research. I participated in forming collaborations with other entities, initiating new projects and mentoring students. Most of my work revolves around our traffic simulator, CityMoS, where I design models that depict commuter behaviour in different scenarios thus bringing more realism into our simulator and providing better accuracy for decision makers.

 

In 2017, I also became part of the Cooling Singapore project, providing my expertise in mobility-related modelling. I got a chance to work together with fellow researchers and professors from MIT (Massachusetts Institute of Technology), ETH Zurich (Swiss Federal Institute of Technology in Zurich) and NUS (National University of Singapore). I believe that the project was a great way to do something meaningful for Singapore.

 

One of the best aspects of working in CREATE is that you are surrounded with people from world-renowned universities. I immensely enjoyed working with Prof. Alberto Sangiovanni-Vincentelli from Berkeley on our common project, BEHAVE, that dealt with studying the interactions between humans and autonomous vehicles, which was within the first batch of funded seed projects in CREATE.

 

Q: What are some of your plans in the future?

I would love to see the research that we have done for so many years turn into a product that is being used by companies and city planners to aid their decision-making process when it comes to mobility. I would love to invest time and effort in making this happen. Apart from that, I think that I am still at an age where I need to actively learn new skills in order to enrich my perception of the world, the problems that our society faces and the possible solutions. I would, therefore, be looking for opportunities that would allow me to do that.

 

I have really enjoyed teaching during the last two years. As I mentioned I was lucky to have some very good teachers throughout my education and I would very much like to pass on this kindness and be a good teacher to my students. I am still quite a novice at teaching and I believe I have a lot more to learn. I hope I will have a chance to continue with this endeavor in the future.

 

(We asked about a fictional character in our last interview, who would be a ‘dream’ teammate in real-life for Jordan?)

 

Q: If you could have anyone in the world, past or present, as your research teammate, who would you choose and why?

To me, what makes the difference between a good and a great team in research is interdisciplinarity. Therefore, I would be interested in working with a researcher from a different field than mine. Even though there are many exciting disciplines out there, I would probably go for a scientist from the field of cognitive neuroscience or cognitive psychology. The reason for this is that I see the human brain as a very skilled modelling machine. In a way, we all have a running model of the surrounding world inside our minds. It is a model that we constantly utilize to perform extremely complex tasks and decisions and I believe that modellers can learn a lot from how the brain creates, updates and maintains this model. Ironically, and unfortunately, we do not really have access to the inner workings of our minds so we need the fields of cognitive neuroscience and cognitive psychology, which are rapidly progressing, to help us peek into our “natural modelling machines”.

 

There are many scientists that are doing groundbreaking research in those fields but if I have to choose just one, I would go in the direction of Prof. Geraint Rees from my alma mater, UCL. I am particularly interested in the concept of attention, which is one of his core research interests, and it is getting increasingly popular in the field of machine learning as well.

 

_____________________________________

 

This interview definitely provided us a sneak peek into Jordan’s mind, or should we call it his ‘natural modelling machine’? His contributions in CityMoS, the Cooling Singapore project and BEHAVE are true testaments to his expertise and we are glad that the little misunderstanding brought him here!