ELM-FBPINNs: an efficient multilevel random feature method.
Significantly accelerated the training of physics-informed neural networks by combining them with multiple levels of domain decompositions and extreme learning machines.
Anderson, S., Dolean, V., Moseley, B., Pestana, J. (2026).
Machine Learning for Computational Science and Engineering.
Paper    Code

Local feature filtering for scalable and well-conditioned domain-decomposed random feature methods.
Significantly accelerated the training of physics-informed neural networks by using random features and domain decomposition to turn their optimisation problem into a structured least squares problem and proposing a novel preconditioner to accelerate convergence.
van Beek, J. W., Dolean, V., Moseley, B. (2026).
Computer Methods in Applied Mechanics and Engineering.
Paper    Code    Workshop

A differentiable hybrid modeling approach for learning soil water retention mechanisms from partial knowledge and data.
Reframed how soil water retention is modelled by augmenting traditional modelling with machine learning components that learnt previously uncertain physics directly from data. Using this method, we accurately reproduced soil water retention behaviour and uncovered new pore-scale insights.
Norouzi, S., Moldrup, P., Moseley, B., Robinson, D., Or, D., Hohenbrink, T. L., Minasny, B., Sadeghi, M., Arthur, E., Tuller, M., Greve, M. H., de Jonge, L. W. (2026).
Journal of Hydrology.
Paper    Imperial News Article

Learning, Solving and Optimizing PDEs with TensorGalerkin: an efficient high-performance Galerkin assembly algorithm.
Accelerated the solving, learning, and optimization of complex physical systems by reformulating finite element PDE computations as GPU-efficient tensor operations that eliminated costly loops, enabling faster, more scalable, and more accurate performance across simulation, inverse design, and physics-informed learning tasks.
Wen, S., Chi, M., Yu, T., Moseley, B., Michelis, M. Y., Ren, P., Sun, H., Mishra, S. (2026).
Proceedings of the 43rd International Conference on Machine Learning.
Paper

Modern, Efficient, and Differentiable Transport Equation Models Using JAX: Applications to Population Balance Equations.
Developed a modern JAX-based solver that achieves up to 300x acceleration in population balance equation simulations, enabling faster, fully differentiable models to automate engineering processes and shorten drug development timelines.
Alsubeihi, M., Jessop, A., Moseley, B., Fonte, C. P., Rajagopalan, A. K. (2025).
Industrial & Engineering Chemistry Research.
Paper

History-Matching of imbibition flow in fractured porous media Using Physics-Informed Neural Networks (PINNs).
Demonstrated that physics-informed neural networks can efficiently and accurately model multiphase flow in fractured porous media, validated against experimental data and numerical simulations.
Abbasi, J., Moseley, B., Kurotori, T., Jagtap, A. D., Kovscek, A. R., Hiorth, A., Østebø Andersen, P. (2025).
Computer Methods in Applied Mechanics and Engineering.
Paper

Challenges and advancements in modeling shock fronts with physics-informed neural networks: A review and benchmarking study.
Explores how to improve physics-informed neural networks for solving equations that involve sudden physical changes – like shock waves – highlighting their current limitations and the need for better techniques to handle complex problems accurately.
Abbasi, J., Jagtap, A. D., Moseley, B., Hiorth, A., Andersen, P. Ø. (2025).
Neurocomputing.
Paper

Multilevel domain decomposition-based architectures for physics-informed neural networks.
Improved the performance of physics-informed neural networks by using multiple levels of domain decompositions to model different frequency scales in their outputs.
Dolean, V., Heinlein, A., Mishra, S., Moseley, B. (2024).
Computer Methods in Applied Mechanics and Engineering.
Paper    Code

Finite basis physics-informed neural networks (FBPINNs): a scalable domain decomposition approach for solving differential equations.
Proposed a new method for solving differential equations which is able to scale to large problems by using physics-informed neural networks and a divide-and-conquer strategy.
Moseley, B., Markham, A., Nissen-Meyer, T. (2023).
Advances in Computational Mathematics.
Paper    Code

Super Resolution of Lunar Satellite Images for Enhanced Robotic Traverse Planning: Maximizing the Value of Existing Data Products for Space Robotics.
Used deep learning to super-resolve images of the lunar surface by 2x, allowing us to see into small craters with higher detail.
Delgado-Centeno, J. I., Harder, P., Bickel, V., Moseley, B., Kalaitzis, F., Ganju, S., Olivares-Mendez, M. A. (2023).
IEEE Robotics & Automation Magazine.
Paper

Seismic savanna: machine learning for classifying wildlife and behaviours using ground-based vibration field recordings.
Showed machine learning can identify elephants and other animals using the seismic vibrations they make.
Szenicer, A., Reinwald, M., Moseley, B., Nissen-Meyer, T., Mutinda Muteti, Z., Oduor, S., McDermott-Roberts, A., Baydin, A. G., Mortimer, B. (2022).
Remote Sensing in Ecology and Conservation.
Paper    Code

Cryogeomorphic Characterization of Shadowed Regions in the Artemis Exploration Zone.
Used deep learning to enhance satellite images of proposed landing sites on the Moon for NASA's crewed Artemis mission, in particular those which contain permanently shadowed regions, allowing us to see into them with unprecedented detail for the first time.
Bickel, V. T., Moseley, B., Hauber, E., Shirley, M., Williams, J.-P., Kring, D. A. (2022).
Geophysical Research Letters.
Paper    All News Articles

Seismic localization of elephant rumbles as a monitoring approach.
Showed that the location of elephants can be accurately determined by using the ground vibrations they make whilst calling each other and a positioning algorithm which is able to say how uncertain it is.
Reinwald, M., Moseley, B., Szenicer, A., Nissen-Meyer, T., Oduor, S., Vollrath, F., Markham, A., Mortimer, B. (2021).
Journal of The Royal Society Interface.
Paper    Code    Oxford News Article    Blog Post

Peering into lunar permanently shadowed regions with deep learning.
Used deep learning to enhance satellite images of many different permanently shadowed regions on the Moon, allowing us to see features within these regions down to 3 m in size for the first time.
Bickel, V. T., Moseley, B., Lopez-Francos, I., Shirley, M. (2021).
Nature Communications.
Paper    Oxford News Article    Imperial News Article

Extreme Low-Light Environment-Driven Image Denoising Over Permanently Shadowed Lunar Regions With a Physical Noise Model.
Used a combination of physical modelling and deep learning to enhance noisy satellite images of permanently shadowed regions on the Moon.
Moseley, B., Bickel, V., Lopez-Francos, I. G., Rana, L. (2021).
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
Paper    Blog Post

Deep learning for fast simulation of seismic waves in complex media.
Discussed in detail the advantages and challenges of using deep learning for simulating seismic waves, and presented multiple deep neural networks which can significantly accelerate traditional methods.
Moseley, B., Nissen-Meyer, T., Markham, A. (2020).
Solid Earth.
Paper    Code

Unsupervised Learning for Thermophysical Analysis on the Lunar Surface.
Used machine learning to disentangle the underlying factors of variation in the Moon's surface temperature, allowing us to identify thermal anomalies on the surface and verify existing physical models.
Moseley, B., Bickel, V., Burelbach, J., Relatores, N. (2020).
The Planetary Science Journal.
Paper    Intel News Article    Blog Post

Differentiable Hybrid Modelling for Learning and Optimising Chemical Transport Processes from Experimental Data.
Improved how accurately crystallisation processes can be modelled and controlled, by embedding neural networks inside a differentiable, GPU-accelerated physics solver that learns unknown rate laws and corrects uncertain initial conditions directly from experiments. This makes fast, large-scale optimisation of chemical manufacturing processes possible, such as tuning temperature profiles to shape crystal morphology.
Jessop, A., Alsubeihi, M., Moseley, B., Rajagopalan, A. K. (2026).
arXiv.
Paper

LunarFM: A Shared Multimodal Representation of the Moon's Surface.
Enabled faster, data-driven exploration of the Moon's surface and its resources, by training a foundation model that fuses observations from six instruments across three lunar missions into a single shared representation, letting scientists search for geologically similar regions, map minerals and classify terrain from very few labels, which could help guide future landing-site and resource-prospecting decisions.
Girona-Mata, M., Gawlikowski, J., Goski, S., de Fourtou, G. B., Bickel, V. T., Moseley, B., Calzada-Diaz, A., Kaczmarek, S., Ramos-Pollán, R. (2026).
arXiv.
Paper

PTL-PINNs: Perturbation-Guided Transfer Learning with Physics-Informed Neural Networks for Nonlinear Systems.
Enhanced physics-informed neural networks by combining classical perturbation theory with transfer learning so they can solve weakly nonlinear differential equations more accurately and up to an order of magnitude faster than traditional numerical solvers.
Alexandrino, D., Moseley, B., Protopapas, P. (2026).
arXiv.
Paper

Solving the wave equation with physics-informed deep learning.
Showed that physics-informed neural networks can be used to simulate complex wave physics, and potentially offer a faster approach than traditional simulation methods.
Moseley, B., Markham, A., Nissen-Meyer, T. (2020).
arXiv.
Paper

Fast approximate simulation of seismic waves with deep learning.
Showed that deep neural networks can accelerate the simulation of seismic waves by multiple orders of magnitude compared to traditional simulation techniques.
Moseley, B., Markham, A., Nissen-Meyer, T. (2018).
arXiv.
Paper    Code

Bayesian optimisation for variational quantum eigensolvers.
Showed that optimisation algorithms that are uncertain about their inputs can effectively combat noise in quantum computers, allowing them to accurately and efficiently compute the ground state energy of molecules.
Moseley, B., Osborne, M., Benjamin, S. (2018).
Paper    Code

Physics-informed machine learning: from concepts to real-world applications.
Investigated and discussed the challenges of scaling physics-informed machine learning (PIML) algorithms so that they can solve complex, real-world problems. Designed multiple different PIML algorithms for carrying out seismic simulation, low-light image enhancement, and thermal anomaly detection on the Moon.
Moseley, B. (2022).
University of Oxford.
Thesis