Image generated by Flux Dev 

project summary: 

In early design stages, architects need to address several requirements pertaining to the movement of occupants throughout the building, including the critical case of emergency evacuation. To ensure the safety of users, governing authorities outline in building codes design requirements for means of egress. Building codes take into consideration building function, number of occupants, and user mobility when setting performance requirements, which in turn makes these rules quite complex and challenging to implement. Predominantly, architects and building authorities check compliance manually, which is a time-consuming and error-prone process. In research and industry, several attempts have been made to automate the compliance checking process.  

Automating code compliance checking for egress in particular faces two main challenges. The first challenge is identifying possible escape routes of occupants and checking their properties for compliance, and the second challenge lies in translating building code rules and regulations written in natural language into computable formats. This research aims to introduce a hybrid approach that integrates algorithmic tools for evaluating circulation in early-stage spatial configurations with AI methods to aid architects in refining layouts and producing alternative solutions.