This has been a joint collaboration with the Universities of Arizona and Nottingham. It is one of the first implementations of very large neighbourhood structures. We extended the idea of large neighbourhoods to real instances of timetabling problems with a complex and interacting set of constraints. Discovery of the sequential moves used in our neighbourhood structure was from our innovative visual interface for timetabling problems (STARK). This interface has come out of our research with cognitive scientists at Nottingham University as part of an EPSRC funded project on humanising automated scheduling systems.