About

noodl physics is based on a simple idea: many physical systems that appear very different share the same underlying network structure. noodl physics provides a general, differentiable framework for modelling such systems, allowing the same models to be used for simulation, inference and optimisation.

The idea grew out of discussions between John Craske and Maarten van Reeuwijk at Imperial College London, starting around 2020, about complex physical networks and the surprising amount of mathematical structure shared by models from different disciplines. noodl physics is an attempt to turn that observation into a practical modelling framework: small enough to understand, general enough to extend, and flexible enough to connect different physical systems.

For several years, noodl physics remained largely an idea. The emergence of AI-assisted software development changed what was practical. Using Claude made it possible to turn a concept we had discussed for years into a working framework in a remarkably short time.