Project Summary
Automated real-time process to select the best model structure
Project Achievements
The project focused on three main areas of work: 1) research into strategies to radically speed up resilience modelling in strategic and real-time environments; 2) research into resilience datasets and methods to link these to existing strategic modelling assets; and 3) development of a system to demonstrate the benefits of the approach.
Conclusions
The project has delivered a tool that enables users to quickly simulate the impact of flooding on major transport networks. The system scales to some of the UK’s largest road models, delivering results to a web application in a fraction of the time of alternatives. This provides immediate insights for transport planners and network managers alike.
The project has also provided initial research into automated network development and network pruning techniques to enable much faster runs within operational environments. The project has identified critical complexities in developing real-time ‘computationally rational’ systems which still need to be overcome, but created capabilities and code that we expect to have significant long-term efficiency benefits on wider transport modelling.
Both elements of the project are being taken forward. The resilience simulation will form part of the Cadence ‘What-If’ suite of tools. Additional resilience datasources have been identified which will enhance this toolset in the future. The fundamental research is proposed to be progressed through Innovate UK funding with the computer science department of a leading UK university.
