Skip to main content

Machine learning for physical systems

Research Lab

Developing machine learning approaches for complex physical systems, from structural dynamics and health monitoring to scientific computing.

View research
  • Structural dynamics
  • Scientific machine learning
  • Structural health monitoring
  • Fluid–structure interaction

Research direction

From physical behaviour to faster engineering insight.

A public, concept-level view of how the Research Lab connects engineering systems, computation, and machine learning.

Illustrative structural surface responding to a changing load

01 / 03

Physical system

Engineering structures respond to changing forces over time. The starting point is the physical behaviour that needs to be understood.

  • Physical systems
  • Structural dynamics
Illustrative finite-element mesh showing a structural response

02 / 03

Numerical simulation

Numerical models provide a computational view of that response, creating a bridge between engineering questions and repeatable analysis.

  • Numerical simulation
  • Scientific computing
Illustrative network connecting a learned representation to a response curve

03 / 03

Surrogate-enabled direction

The longer-term direction connects simulation and machine learning to support faster repeated engineering analysis.

  • Machine learning
  • Engineering workflows