Alessio Quaglino is a young mathematical engineer who has a passion for applying his analytical skills in contexts that are apparently very distant, such as Formula 1 racing and cardiology.
What brought you to USI?
In 2008, I graduated in engineering mathematics at the Polytechnic University of Milan, but I also have a second degree, obtained the year before, in engineering physics from the KTH Royal Institute of Technology in Stockholm, Sweden – basically, with only one additional year of studies, I managed to obtain a double degree. I then enrolled at the University of Göttingen, in Germany, where in 2012 I obtained a Ph.D. in applied mathematics. After that, I relocated to the U.K, where I spent three years working for the McLaren F1 team based in Woking.
I knew of a classmate from the Polytechnic who was doing his doctorate at USI, and thanks to him, I learned about the research that was being conducted in Lugano. At that time, I was considering a move back into the academic sector, and I did not want to return to Germany because the overall conditions did not suit me. I have always been attracted by Switzerland, as a country and for its working conditions. Eventually, in 2015, I joined the Institute of Computational Science, a research unit of the USI Faculty of Informatics.
Tell us more about your experience in Formula 1
I landed the job at McLaren quite by chance, really. Let me start by saying that I have always been attracted to the world of motor sports (I was involved in go-kart races until my early twenties). I simply sent my CV without great expectations. Then, unexpectedly, they contacted me: the team was seeking a vehicle dynamics engineer, someone who could produce mathematical models to study the dynamic behavior of suspensions. I was to work on the development of new suspension prototypes, in the form of mathematical concepts for the in-house simulator on which the drivers would carry out driveability tests, and then give us their feedback. If the driveability improved, then we would proceed with the development and try to figure out how to realise the original theoretic idea. I would follow the project up until the telemetry analysis phase, on the racing track.
So, mathematics applied to automotive engineering…
Right, though my job had more to do with mathematics and concepts rather than pure engineering. In fact, when it came to developing the real suspensions, my colleagues and I would interact with the engineers in charge of designing and building the parts.
Would you be at the races as well?
I would be on the racing tracks for the off-season tests, in winter, whereas during the championship races I would assist the free practice sessions, on Fridays, by radio from the headquarters in Woking. However, sometimes we would receive requests from the pit lane to look into specific problems that the drivers would encounter. Normally, this would occur on Fridays, also because no further modifications were allowed after the qualifications, on Saturdays.
Let us talk now about your current activity at USI
I consider my role rather transversal, in comparison to my colleagues. Now I am working on a research project at the Center for Computational Medicine in Cardiology (CCMC), where I deal with simulations for the heart, helping doctors to diagnose and experiment therapies in a virtual environment. My contribution, and that of my colleagues, consists in using real data from patients to create a virtual heart. This is very important to make diagnoses, for example by altering physical parameters (e.g. cell chemistry or muscle fibers) to virtually recreate the disease by understanding the real causes, and to test possible therapies, by applying them to the model to see if they act properly. All this is very similar to what happens in Formula 1, where at first you try the parameters that best explain the data measured by telemetry and then use these values as a starting point from which to improve and optimize the car design.
So, you are active also in the clinical area?
My work consists mainly in developing mathematical-computer tools to help physicians and biomedical engineers in the planning aspects that occur before surgery in the operating room, starting from the evaluation of the actual need to operate or not. For instance, this is very useful when suggested therapies fail to work on certain patients and are successful on others who, however, apparently shared the same problem. When the reasons for this are unknown, that is when I “operate”, with simulation that allow to recreate the pathologies and to apply changes at the microscopic level, down to the single cells. This way the alterations allow to identify the variables that often have an impact on the effectiveness of a therapy or not.
The heart is a biologic organ, what does it have to do with mathematical engineering?
At the end of the day, simulations in cardiology are the same as in Formula 1. It all comes down to understanding which equation can describe a physical phenomenon, and then solving the equation. The functioning of the heart is the result of a chemical interaction at the cellular level that generates electrical pulses, which in turn and on a larger scale, generate a muscular contraction, thus pumping blood into the circulatory system. Whether small or large, these dynamics can be described with physical equations, so really the behaviour of the heart is the result of an interaction between complex systems, just like with Formula 1 vehicles where electronics, thermodynamics and fluid dynamics interact.
What are your future projects?
Generally speaking, I am focusing on a field of research called Uncertainty Quantification, on which I organised a winter school at USI with over 50 participants from all over the world. I became familiar with UQ while I was working in Formula 1 and it was one of the reasons that led me to return to the academic sector to study it more closely. In fact, what I learned in Formula 1 was that when you are doing simulations, often you do not know the parameters as precisely as you would like, so you do not know if you can trust the data, and you do not know how the results will change if you alter the parameters. Therefore, it is very important to treat systematically these parameters and determine how sensitive the data is to uncertainties and how much you can rely on it. Nowadays, the overall approach to simulating a given set of parameters is well established, especially in the field of engineering. Quantifying uncertainties in a systematic way, however, means using a quantitative approach to answer questions, like, "I wonder what would happen if ...".
Alongside my research, together with professor Rolf Krause (director of the Institute of Computational Science at USI), I am part of a start-up company, a spin-off entity called Algo4U, which is based on the idea of developing 'custom' algorithms. In other words, companies that develop their own software programs or systems – e.g. in the medical or mechanical sectors – and who want to innovate their products with more sophisticated algorithms, are normally obliged to acquire skills that are totally unknown to them, which means taking risks and, ultimately, incurring in additional costs. The idea of Algo4U is that these companies can outsource their needs to us, for the mathematical-algorithmic part, and we take care of delivering a software that can improve their products and systems. This method can be applied to any company, in any sector. Ultimately, we do not only provide the maths, but we help also to develop their ideas.
(interview originally published in Ticino Welcome magazine, June-August 2017, p.22-23)