Stronger together with smart algorithms

Peter Bosman (49) grew up in Utrecht and now lives with his wife in Lelystad. After years of renovation work, they enjoy the space and the stunning view their home affords. “I was always one big ball of energy,” he recalls when asked about his childhood. “In primary school I already knew I wanted to work with computers. That was rather unusual at the time, since the personal computer was only just becoming more common. My father, who taught electrical engineering, once brought computers home. I was instantly fascinated, and before I knew it I was programming at the age of ten. Even in secondary school I had just one path, one goal: to study computer science.” With an outstanding academic record (“I scored two perfect tens, even higher than my maths teacher – I don’t think I’ve ever told anyone that...”), he went on to Utrecht University.

A Small Music Studio

The creative side of computer science has always appealed to him. “I have a strong analytical side, but also a creative streak, and computer science is the perfect blend of both. You can come up with new ideas and turn them into something that works, something useful.” Does his creativity extend to other fields? “I love making music, I play keyboard and synthesiser, and I have a small studio at home. From there I collaborate with singers around the world – although these days I don’t have as much time for it, unfortunately.”

His fascination with computer science led to an impressive career. Today Bosman is a group leader at the Centrum Wiskunde & Informatica (CWI) in Amsterdam and a part-time professor at Delft University of Technology. There he works on developing algorithms – the invisible engines behind much of modern technology. His focus is on algorithms in the field of AI.

New medical treatments

So what exactly is an algorithm? “The easiest way to see it is as a set of instructions that a computer follows to perform a task,” he explains. “Just as you follow a recipe to cook a dish, a computer follows an algorithm to solve a problem.” Algorithms are everywhere – in your smartphone, your navigation app, you name it. Bosman specialises in evolutionary algorithms. “They are comparable to natural evolution. Instead of refining a single solution, these algorithms test different combinations of many solutions. They learn to improve themselves by discovering which combinations are more or less useful. With each cycle, they become smarter.”

This approach is powerful when tackling complex issues, especially those involving many factors with no single optimal answer – only trade-offs between them. “That makes them useful, for instance, in planning extensions to the power grid – determining the most future-proof expansion with the most capacity at the lowest cost – or in designing medical treatment plans – deciding which plan offers the best chance of recovery with the fewest side effects.”

Less burdensome

One of the most remarkable applications of his work is in healthcare, particularly in the treatment of prostate and cervical cancer. Bosman collaborates with doctors, physicists, radiation therapy technicians, and researchers at Amsterdam UMC and Leiden UMC on a technique called brachytherapy. This involves placing a small radioactive source close to the tumour through hollow needles to irradiate the cancer from within. The aim: maximum effectiveness with minimal damage to surrounding tissue. “Our algorithms help determine exactly where and for how long the source should pause in the different needles to strike the right balance: delivering enough radiation to the tumour while sparing healthy tissue as much as possible. That makes treatment more precise, safer, and less burdensome for the patient,” he explains.

A day to remember

He conducts this research together with his wife, Tanja Alderliesten. She works in the radiotherapy department at Leiden UMC and specialises in medical informatics – a perfect bridge between the technical and the medical sides of algorithms. It was their idea to improve the process of designing brachytherapy treatment plans. Today they still work side by side on further developing this application. Bosman recalls: “On 17 March 2020 – the exact day two years after my father passed away from cancer – the world was in lockdown, everything seemed to stand still. But that day we watched as the first prostate cancer patient at Amsterdam UMC was treated using a plan generated by our algorithm. It went perfectly. An unforgettable moment.”

Improving Patients’ Lives

The project is a textbook example of how fundamental research (developing new methods) and applied research (direct use in practice) can go hand in hand. “That’s what I love most,” Bosman says. “You create new knowledge, and at the same time you see that knowledge directly improve something – in this case, patients’ lives and the work of medical experts.”

The applications of his work go beyond medicine. Bosman and his team have also applied their algorithms to logistics challenges and optimising power grids. In a new project, they are joining forces with partners DIFFER and neighbouring institute AMOLF to automatically design the solar cell of the future. Wherever smart, insightful decision-making is needed, his algorithms can play a role.

A tool

In an era when AI is advancing rapidly and sometimes causing concern, Bosman takes a measured view. “AI is a tool. It is not inherently good or bad. What matters is how we use it. Transparency and control are essential.” He acknowledges the risks – lack of explainability or unfair outcomes. “That is why Tanja and I also work on ‘explainable AI’: AI that is understandable and verifiable, so we can see how a recommendation is reached.”

A dream in progress

Bosman is building AI that genuinely supports humans in making complex decisions. The evolutionary algorithms and neural networks – both forms of AI – he develops are not intended to replace people, but to serve as tools that make processes faster, safer, and more insightful. His ambition is to expand this approach further: to see AI not just as a technical aid, but as a way of working together to advance science, healthcare, and society. A dream that will never truly be finished.

Could he have predicted as a child that one day he would be collaborating with doctors and scientists on such applications? Probably not. But the drive to understand and to make possible what once seemed impossible – that was always there. And it has never gone away.

This article was published in the Amsterdam neighbourhood newspaper DWARS (reproduced with permission).

Text: Anita Boelsums

Header picture: Frank Schoevaart