Modern medicine is both blessed and cursed with vast amounts of data over wide ranges of spatial and temporal scales. Understanding how these data relate requires multiscale modelling to link such measurements via their biochemical, biophysical, anatomical, and developmental processes.In order to share this understanding, we must create precise descriptions of models, enabling them to be interpreted in ways that minimise the introduction of errors and ambiguities.The aim of the FieldML project is to define a common format for describing the representation of, and operations on, scalar, vector, and tensor fields used in computational models of biological systems.Work on the development of FieldML has been the major focus of my James Cook Research Fellowship. In this talk I will report on the aims, challenges, and progress with FieldML.