Centrum Wiskunde & Informatica

Mathematics, simulation and AI can help industry address complex challenges – from forecasting cable temperatures and simulating urban environments to improving R&D decisions and engineering design. On 29 September, the symposium 𝗥𝗲𝗰𝗲𝗻𝘁 𝗔𝗱𝘃𝗮𝗻𝗰𝗲𝘀 𝗶𝗻 𝗜𝗻𝗱𝘂𝘀𝘁𝗿𝗶𝗮𝗹 𝗮𝗻𝗱 𝗔𝗽𝗽𝗹𝗶𝗲𝗱 𝗠𝗮𝘁𝗵𝗲𝗺𝗮𝘁𝗶𝗰𝘀 will bring together researchers, companies, research institutes and policymakers at the JADS - Jheronimus Academy of Data Science in ’s-Hertogenbosch. The programme combines keynote lectures, presentations by academic–industrial duos, interactive discussions and networking. A panel will explore the future role of mathematics, simulation and AI in engineering and industrial innovation. Among the speakers are CWI researchers Peter Grunwald, who will deliver a keynote lecture, and Dimitris Loukrezis, who will present together with Kujtim Gashi of QUGATE. The symposium also looks ahead to ICIAM 2027, the world’s leading congress on industrial and applied mathematics, which will take place in The Hague next year. Interested in what mathematics can contribute to technological development and industrial practice? View the programme and register by 21 September 👉 https://lnkd.in/efbFfTPe #AppliedMathematics #Industry #Engineering #AI 𝘐𝘮𝘢𝘨𝘦 𝘨𝘦𝘯𝘦𝘳𝘢𝘵𝘦𝘥 𝘸𝘪𝘵𝘩 𝘈𝘐

The Netherlands’ first programmable computer barely worked. To mark CWI’s 80th anniversary, we are revisiting images from our history and the people behind them. This time: 𝗔𝗥𝗥𝗔. Its name stood for Automatische Relais Rekenmachine Amsterdam (Amsterdam Automatic Relay Calculator). On 21 June 1952, Theo Rutten, then Minister of Education, Arts and Sciences, officially put ARRA into operation at the Mathematical Centre, now CWI. At the press of a button, Rutten launched a program that generated a sequence of random numbers – a considerable feat at the time. But according to its builders, Carel Scholten and Bram Loopstra, the machine achieved little after its official debut. 🗨️ ‘After the opening, as far as I can remember, it never produced another meaningful line of output,’ Scholten later recalled. ARRA was plagued by reliability problems arising from several parts of its design. One persistent source of trouble was its approximately 1,200 relays. Their switching times were unreliable and their contacts deteriorated. The relays were later covered with small plastic bags to protect them from dirt, but the problems continued. The contacts had to be cleaned and the switching times readjusted regularly – a recurring chore known as 𝘱𝘰𝘰𝘵𝘫𝘦𝘴 𝘱𝘰𝘦𝘵𝘴𝘦𝘯. The builders decided to start again. Eighteen months later, Scholten and Loopstra, joined by computer pioneer Gerrit Blaauw, completed ARRA II. Despite its name, it was not a repaired version of ARRA but an almost entirely new machine. The unreliable relays – switches with moving parts – had been replaced by electronic components, and faulty parts could be exchanged more easily. ARRA II still had some technical problems, but it was far more reliable and successfully performed useful calculations. In 1956, computer construction moved into industry with the establishment of Electrologica, CWI’s first spin-off company. At the Mathematical Centre, attention increasingly shifted towards programming languages and algorithms. 𝗘𝗱𝘀𝗴𝗲𝗿 𝗗𝗶𝗷𝗸𝘀𝘁𝗿𝗮 became one of the pioneers of this new direction. His shortest-path algorithm still underpins route planning in many modern navigation systems. We will tell his story in our next history post. Photo 1: Dineke Botterweg, one of the 𝘳𝘦𝘬𝘦𝘯𝘢𝘢𝘳𝘴𝘵𝘦𝘳𝘴 (the women who performed lengthy calculations at the MC), operates ARRA, 1952. Photo 2: Minister Theo Rutten presses the button to inaugurate ARRA. Builder Bram Loopstra stands on the left; on the right is Mayor of Amsterdam Arnold d’Ailly, 1952. Photo 3: Some of ARRA’s approximately 1,200 relays, 1952. Photo 4: ARRA II at the Mathematical Centre, 1953. CWI Archive. #cwi80 #computinghistory

Before CWI had a computer, it had 𝘳𝘦𝘬𝘦𝘯𝘢𝘢𝘳𝘴𝘵𝘦𝘳𝘴: women who performed complex calculations by hand. To mark CWI’s 80th anniversary, we are sharing images from our history, and the stories of the people behind them. We begin with a group whose contribution has often been overlooked. When CWI was founded as the Mathematical Centre in 1946 to support post-war reconstruction in the Netherlands, a team of women carried out lengthy and complex calculations using mechanical desk calculators. Work that could take them months can now be completed in a fraction of a second. They later learned to operate and program the ARRA, the first Dutch computer, an enormous machine that occupied much of a room. Former rekenaarster Dineke Botterweg recalled it as ‘incredibly stupid’ in Dutch newspaper de Volkskrant in 2024: ‘If even one plug was in the wrong place, it would stop working.’ The women used the ARRA for calculations related to the wings of Fokker aircraft and to storms and spring tides for the Delta Works. By telling the machine exactly which steps to perform, they were doing what we would now call programming. In the decades that followed, women became increasingly underrepresented in computing, while the pioneering contribution of the rekenaarsters largely faded from view. Eighty years later, their work deserves a prominent place in CWI’s history. 𝗡𝗲𝘅𝘁 𝗶𝗻 𝘁𝗵𝗶𝘀 𝘀𝗲𝗿𝗶𝗲𝘀: ARRA, the first Dutch computer. Photo 1: 3 𝘳𝘦𝘬𝘦𝘯𝘢𝘢𝘳𝘴𝘵𝘦𝘳𝘴 at the Mathematical Center in 1954. Photographed by J.A. Zonneveld. Photos 2 and 3: part of the 𝘳𝘦𝘬𝘦𝘯𝘢𝘢𝘳𝘴𝘵𝘦𝘳𝘴 team in 1948. CWI archive. #cwi80 #computinghistory

Digital security needs to be ready for the quantum era 🔐. Researcher Wessel van Woerden has received a Veni grant from NWO (Nederlandse Organisatie voor Wetenschappelijk Onderzoek) to develop compact digital ‘locks’ that can withstand attacks by quantum computers ⚛️. Current post-quantum security solutions are often too large to replace existing cryptographic systems without causing delays or technical complications. Using advanced geometry, Wessel will investigate mathematical structures that can encode data more efficiently while remaining secure. He will join CWI’s Cryptology group in January to carry out the research. Congratulations, Wessel! 👏

What if researchers could continuously monitor statistical evidence as it accumulates? Rather than having to wait until a study has finished before drawing conclusions, such an approach would allow evidence to be assessed as it emerges - for example, in rapidly evolving situations such as a pandemic. This is exactly the idea behind e-values, an emerging framework for statistical testing with applications such as clinical trials, A/B testing, political polling and financial risk management. CWI researchers played an important part in the development of this highly international field. This month, ERCIM News devotes a special issue to e-values. Guest-edited by Peter Grunwald Grünwald (CWI & Leiden University), Wouter Koolen (CWI & University of Twente) and Johanna Ziegel (ETH Zürich), it opens with a keynote by Michael I. Jordan (Inria). This special theme includes ten contributions by researchers in the Netherlands, alongside contributions written by researchers from, among others, Inria, ETH Zürich, Carnegie Mellon University, Microsoft Research and Netflix. Recommended reading for anyone interested in the future of statistical inference. Congratulations to the guest editors and all authors on this special issue! 🎉

It is with great sadness that we have learnt of the loss of Frank Nack. Frank was a postdoctoral researcher at CWI for six years, from 2000 to 2006. He will be greatly missed by everyone at CWI who knew him. Our sincere condolences to Frank’s family, friends, and colleagues.

Missed some of the latest CWI stories? Our latest newsletter brings together recent research news, media highlights and upcoming events in one concise overview. Read the latest edition and subscribe for the next one!

🏆 Huge congratulations to CWI researcher Jurgen Vinju and his co-author Terence Parr (the creator of ANTLR) for receiving the prestigious 10-Year Most Influential Paper (MIP) Award at the SLE Conference (Software Language Engineering) in Rennes! 🎓✨ They are being honored for their groundbreaking 2016 paper, "Towards a Universal Code Formatter through Machine Learning". A decade ago, Jurgen and Terence introduced CODEBUFF, a revolutionary tool that changed how we think about code formatting (making code clean, structured, and readable for humans without changing how the software actually works). Instead of painstakingly writing hundreds of complex, language-specific rules by hand, CODEBUFF pioneered the use of machine learning to automatically learn and apply formatting styles directly from example repositories. Back then, it was a "drop-the-mic" moment: it proved you couldn't get more elegant with Machine Learning, outperforming the average programmer. Fast forward to 2026, and this 10-year-old paper is more relevant than ever. In our current era of massive Large Language Models (LLMs) and generative AI, code formatting has become an environmental question. Asking power-hungry cloud giants to format code is like using a sledgehammer to crack a nut. As Jurgen highlighted in his talk, looking at this through a 2026 lens means embracing Green Computing. By shifting formatting tasks from giant, power-hungry LLMs to tiny, local AI models (one per language), the software industry might achieve massive global energy savings. This innovative approach continues to inspire and is a prime example of research that stands the test of time. Jurgen presented their honorary talk today, reflecting on a decade of universal code formatting. CWI is incredibly proud to see this work recognized on the global stage! 🥂👏 👉 Read the paper here: https://lnkd.in/efwkiieJ #CWI #SoftwareEngineering #MachineLearning #SLE2026 #MIPAward #ResearchImpact #Innovation #ANTLR #CodeFormatting #ComputerScience #GreenComputing #LightweightAI

What does it take to keep society functioning during a large-scale crisis? CWI is contributing to VORTEX, a new research project led by Vrije Universiteit Amsterdam that has received €541,000 in funding from TKI DINALOG. The project investigates how emergency support points and coordination points can help the Netherlands become more resilient during prolonged disruptions. These points could support citizens with basic services and help align civil and military logistics when systems such as energy, communication or transport are under pressure. CWI contributes expertise in quantitative modelling and optimization, helping to assess what works in different crisis scenarios. Read the full article on our website 🔗 https://edu.nl/h73ky

Congratulations Junqiao Lin!! We’re very proud of you here at Centrum Wiskunde & Informatica!