Casper Van Gheluwe studied Computer Science Engineering at UGent and is now a PhD student within the Industrial Systems Engineering and Product Design research group of his alma mater. We talked to him about his role within TMaaS.
What is the subject of your current research at UGent?
I look at how to set up powerful models for urban mobility via new, crowdsourced data sources. One example of this would be how the control of traffic lights at intersections can be improved using GPS data. In the first instance we look into how current simulation models and programmes work. We also evaluate the quality of the available data sets. This part of the subject is more or less finalised.
The next phase involves creating reliable traffic simulations. That is consistent with the work carried out within TMaaS. Information about traffic flows (of cars, cyclists and pedestrians) is currently obtained in the main from counting elements that are clustered in blocks of 15 minutes. This can produce a distorted picture of the actual traffic situation. Cyclists often arrive in flurries because they were stuck at a red light at a previous intersection. If you follow individual cyclists with GPS data sources, you can adjust the control of the lights at the different intersections accordingly. This improves the flow for cyclists and it could also be applied to vehicles: you can ensure that the traffic lights for cyclists only turn green when a new flurry comes along.
We are currently working with two specific use cases. The first focuses on traffic on part of the R40. We have precise data about the control of the traffic lights and the number of cyclists as well as our own data. In the other use case we are investigating the traffic situation around the campus in Zwijnaarde. We note that there are particularly frequent mobility problems around school times. We can propose interesting solutions based on our simulations.
At first glance, the connection with TMaaS seems to be limited, but in practice we could use these simulations in real time when we see an incident on the dashboard. This way we can make a “digital twin” of a certain intersection or part of the city. Traffic operators can test an adaptation in the virtual world, and then go on to apply it in reality. That is a very interesting avenue for TMaaS to look at.
Can you give us a foretaste of the TMaaS webinar that will take place on 28 May and give us more information about the prototype that you have created?
I will be presenting the webinar together with colleagues from TomTom, Be-Mobile and Waylay. They will present the possibilities, architecture and data sources of TMaaS.
Any cities that are interested will have already received an initial explanation during an earlier webinar. This time we go into greater technical detail so that the cities with a particularly strong interest can discover the possibilities that TMaaS has to offer and which data they would need to provide. It is my job to generate enthusiasm among these cities for the project: what possibilities does the dashboard offer them? What is our current status on the project?
I will also show them prototypes of specific map visualisations of the traffic situation in Ghent, such as graphs that show the evolution of traffic in the car parks in Ghent. The prototype does not yet provide a global picture of the service we can provide, but I will be able to highlight some aspects.
Why have you chosen to work in the academic world and not in the private sector?
After completing my studies, there were two options in the private sector that seemed viable. I could work as a consultant with major IT companies, but then I would always be on the road. That didn’t appeal to me very much. I also had the possibility of choosing a job as a junior developer at a small- or medium-sized company. But I wondered whether this kind of job would be able to offer me enough variety. Moreover, my qualification is very specific and programming is something you can do with a very broad and non-specific range of qualifications. The academic world seemed to offer me more variety and freedom, and more opportunities to express my personal vision.
I was actually drawn into doing a PhD because while I was studying, I did some vacation work with my supervisor, Prof. Sidharta Gautama. After completing the thesis for my Master’s Degree I was asked to continue my research as a PhD student and to work on a collaborative and participative venture at UGent. As a PhD student you get to reflect on the policy of the department, and even of the entire university. What appealed to me was the opportunity to make a difference. As a programmer you have little influence on the strategic choices your company makes. Carrying out the kind of work involved in the project also reflects my personal vision of sustainable mobility. It is extremely gratifying to be able to work on something that you actually believe in yourself.
How does a collaborative venture within an international team work (within UGent)?
Our specific group is made up of a wide range of PhD students and other colleagues with international backgrounds. I really like this mix of cultures and find it very educational, because you discover how other people view problems and the different solutions they find. Our working language is English of course. But there are currently two people from Ecuador in my office as well as one from Colombia – so after a bit of time we end up also learning some Spanish. Sometimes colleagues relocate abroad and we find ourselves having to find a suitable time to hold meetings. If it’s nine o’clock in the morning for us, it is two o’clock in the morning for a colleague in South America. So, it’s not always easy.
What are the most important challenges for you within TMaaS?
I have two important personal challenges. The first is the integration of the data sources of the partners into a neutral framework so that data can be easily added whilst respecting the underlying open or commercial license conditions. One of the conditions is that data from the different parties should not be mutually identifiable and not appear in duplicate. That is a major challenge: incidents must not be reported twice for instance, even though they appear in different data sources. That sounds simpler than it actually is.
My role also involves designing the dashboard so that it is useful AND user-friendly for both target groups, namely the traffic operators and citizens. The two groups have their own needs and expectations, but both want to obtain relevant information simply and speedily.
What are your plans for the future?
After the TMaaS project I would like to complete my PhD. And as a matter of fact, in my research group, new projects are constantly being started up, so perhaps there will be one in the future on which I will be able to collaborate.
And one last question: how do you travel to work?
On my bike! Every day I ride seven kilometres each way, in all kinds of weather. By the time I get to work in the morning, I am completely wide awake. It’s just great.