ALT Changes can be necessary at any stage in an infrastructure or building project. As well as yielding direct and indirect rework costs and programme disruption, the increased material wastage and knock-on effects causing further errors and reworks make the post-project approval and apportionment of change-related costs complex and highly contentious. This presentation will give an overview of our current efforts to bring automation in detecting changes and assessment of their impacts using machine learning, impact models and change visualisation.