Keith started academic life as a theoretical physicist and came to mechanical engineering a little later; as a result, his research interests reflect a leaning towards mathematics and computation. His main focus is on the application of ideas from signal processing and machine learning to problems in structural dynamics including: nonlinear dynamics and system identification, structural health and condition monitoring and uncertainty analysis. Particular interests at the moment include Bayesian methods and population-based structural health monitoring. His initial application domain was aerospace industry, but that has widened to include civil infrastructure and energy, with a particular focus on wind energy systems at the moment