“Like artists, good teachers have a trademark style”

For Michael Unser, the link between art and science recurs in conversation. “My mother was an artist and my father a scientist, so it’s no surprise that both art and science are big parts of my life,” he says. “Why should I have to choose between the two anyway?” As if to hammer home his point, he turns around and motions affectionately toward two paintings on his office wall. “One was painted by my other, the other by my sister.”

Unser, who teaches the “Signals and systems” classes at EPFL, won the distinction of best teacher in the microengineering section for 2022. In his expert view, art and science have much in common. He points to his own field of expertise as a case in point: “Music is a great example of signals in action.” An acoustic signal is a variation in pressure over time, which conveys information. “An MP3 file is a classic example of a discrete signal stored on a computer or smartphone,” says Unser, noting more generally that “math is an exercise in precision, clarity and simplicity, so it’s very much like art in its broadest sense.”

An early sign of things to come?

When it came to choosing a career path, Unser followed his heart. “I’d always loved building things, so engineering seemed like a good option.” Was this an early sign of his long and distinguished teaching career at EPFL? Unser completed both his Master’s and his PhD in electrical engineering at the School before taking a decade-long hiatus to work on bioimaging at the National Institutes of Health in Bethesda, Maryland. And all the while, he continued honing his skills as an amateur classical guitarist. “The instrument has a special place in my life,” he says. “To me, practicing is like yoga. It’s way of relieving stress.” Scientist by day, artist by night.

And stress isn’t something to which Unser is immune: he cites paperwork, fundraising and competition as some of its causes. But after more than 25 years as an academic staff member at EPFL, he’s absolutely certain about one thing: “Not a single professor here is in any hurry to retire, me included!” Unser explains: “There’s something magical about teaching, and about research too.” He adds that his role is made all the more magical by the fact that “our School is one of the top universities in French-speaking Europe, and our students are highly motivated.”

Intuition over rigor

Unser also scores high on the motivation scale – and he needs every bit of that energy because he has the highest teaching workload of any professor in EPFL’s School of Engineering (STI). “The biggest challenge in my field of expertise, especially in the ‘Signals and systems’ classes I teach, stems from the sheer sophistication and complexity of the subject matter,” he explains. Those classes cover the basic principles and methods of signal characterization and linear system analysis and description.

“These classes distill 200 years of mathematical precedent,” adds Unser. So how does he go about conveying the density, complexity and rigor of theories dating back to the time of Joseph Fourier and perfected by Laurent Schwartz without leaving his students feeling daunted, discouraged or defeated? Contrary to expectations, Unser relies on intuition. “I start by getting my students to ‘feel’ their way into the material,” he says. “Often, I’ll ask them to tease out the core concepts – especially the notion of an abstract space of infinite size – using analogies.”Only later does he “dispense the dry, heavy theory.” But, he says, treading the fine line between intuition and theory is “challenging in a subject where rigor is so important.”

Errors with far-reaching implications

Speaking of rigor, Unser mentions that after “doing a bit of digging” he found mistakes in a number of seminal university textbooks: “If the teaching material itself contains errors, this can have far-reaching implications for students’ learning.” For Unser, the situation had to be rectified. So he published a paper in which he redefined the necessary and sufficient condition for the stability of a bounded-input, bounded-output (BIBO) system. “Regrettably, the previously accepted condition was incorrect,” he says.

Could this tireless quest to ground teaching in solid foundations be described as the “Unser method”? The question brings a smile to the professor’s face. “Maybe! If so, I’d be delighted. Like artists, good scientists – and, by extension, good teachers – often have a trademark style.”