Google Scholar · ResearchGate · Research.com · NVA · Profil for industripartnere
Turning heavy mathematics into algorithms fast enough to use
I am Chief Scientist at NORCE and a SIAM Fellow, working on numerical methods, imaging and data analysis. Most of my work sits where a computation is either too slow, too noisy or too much of a black box to be useful in practice.
Four kinds of problem I am usually brought in to solve
Seeing what the sensor cannot show. Denoising, sharpening, segmentation and 3D reconstruction of images and scans, including cases with heavy noise, missing data or a single projection. The aim is to mark out an object and measure it reliably, not only to make the picture look better. Relevant to medical scans, industrial inspection, microscopy and remote sensing.
Inferring what you cannot measure directly. Inverse problems: recovering material properties, sources, permeabilities or geometry from indirect measurements, with methods that hold up when the answer is discontinuous and the data incomplete. Relevant to subsurface characterisation, process monitoring and non-destructive testing.
Simulation that runs in minutes, not days. Fast solvers for partial differential equations — operator splitting, multigrid and domain decomposition — and physics-informed neural networks as surrogates when a full simulation is too slow for the decision at hand.
Machine learning you can explain. Deep networks that come with a mathematical account of why they work, and where physical constraints and known geometry can be built into the architecture. This matters when a black-box model is not acceptable to a regulator or a customer, or when training data are scarce.
What this can do for your business
The approach I am developing now joins classical numerical mathematics with modern machine learning: models that learn from your data but obey the physics, run fast enough for live decisions, and can be audited. That opens problems that were out of reach a few years ago.
- Digital twins that keep pace with the plant. Physics-informed models that update from live sensor data — furnaces, reservoirs, pipelines, power systems, patients.
- Inspection and diagnosis without the bottleneck. Detect, delineate and measure defects, anomalies or lesions in images, video and scans, with a confidence level rather than a bare label.
- Decisions on incomplete data. Recover what your sensors cannot see directly, and know how far to trust the answer.
- AI you can certify. Networks with physical and geometric constraints built in, trained on small datasets, explainable to regulators and customers.
- Design and optimisation at speed. Simulation surrogates that let engineers test hundreds of variants in the time one full run used to take.
- One picture from many sources. Fusion of satellite, drone, AIS and in-situ sensor data for monitoring oceans, energy assets and infrastructure.
So far this has served energy and metals, health, maritime, earth observation and manufacturing. The mathematics does not mind which sector comes next.
Ways to work together
A short feasibility study on your own data, to establish whether the problem is solvable and what accuracy is realistic before you commit a budget. Contract research towards a defined algorithm or prototype. An industry-partner role in Research Council of Norway or Horizon Europe projects, where public funding carries much of the cost. Or co-supervised PhD and postdoc positions on your problem, and training for your engineers.
Send me a short description of what you measure, what you need to know from it and how fast you need the answer. That is usually enough for me to say whether this is worth pursuing — and to be honest if it is not.
Background
PhD in applied mathematics, University of Jyväskylä, 1991. Professor at the University of Bergen (1994–2021), Nanyang Technological University, Singapore (2007–2011), and Chair Professor and Head of Mathematics at Hong Kong Baptist University (2017–2022). Chief Research Scientist and Executive Program Director at COCHE, Hong Kong (2022–2023). Chief Scientist at NORCE since 2023. SIAM Fellow (2026), Feng Kang Prize for Scientific Computing (2009), Nanyang Award for Research Excellence (2011), Humboldt Scholarship (1993). More than 250 publications and over 13 000 citations. Board member of NOBIM, the Norwegian association for image analysis and machine learning.
Editorial boards
Editorial work keeps me current on what is actually working across numerical analysis, imaging and machine learning — well before it reaches textbooks or products.
- SIAM Journal on Numerical Analysis (2024–)
- SIAM Journal on Imaging Sciences (2018–)
- Journal of Mathematical Imaging and Vision (2017–)
- Inverse Problems and Imaging (2008–)
- East Asian Journal on Applied Mathematics (2010–)
- International Journal of Numerical Analysis and Modeling (2004–)
- Mathematical Foundations of Computing (2016–)
- Frontiers in Computer Science (2024–)
- Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization (1997–)
- Advances in Continuous and Discrete Models: Theory and Applications — Editor in Chief (2021–2024)
- Numerical Mathematics: Theory, Methods and Applications — Executive Editor (2005–2021)
Publications and full profile
Google Scholar · ResearchGate · Research.com · NVA · Profile for industry partners
