Masonry arch bridges still form the backbone of the UK's transport infrastructure; approaching 50% of bridge spans on the UK rail and regional highway networks are masonry. However, a number of prominent failures suggest we may be at a tipping point - brought about by a perfect storm of the increasing age of the structures, new traffic loading demands, climate change effects pushing structures to new limits and severely restricted maintenance budgets. To respond to the challenging times ahead there is a need to develop a much greater understanding of how real bridges behave, moving beyond traditional 2D idealisations and identifying the extent to which bridges are capable of 'autogenously healing' under cycling loading. This is important as, currently, bridge engineers faced with a damaged bridge simply do not have the tools needed to make informed assessment decisions, and may needlessly strengthen or demolish a structure even if it could, in reality, be repaired at comparatively modest cost. The goal is to provide those responsible for the management of bridges with a powerful suite of analysis modelling tools and a robust overarching multi-level framework capable of being applied to the diverse population of masonry arch bridges in-service today (i.e. undamaged, damaged and repaired). To achieve this a team of experienced researchers with complementary expertise has been assembled. Medium and large-scale experimental tests will be used to develop and validate analysis tools of different levels of sophistication, with high-level, high-fidelity models, capable of simulating the actual masonry bond and material response, used to calibrate novel intermediate-level and lower-level tools suitable for rapid practical assessment.