Optimising Sun Analysis aims to create a user-friendly application that can quickly be applied by architects or urban designers to test existing buildings and derive new urban forms in a myriad of locations. By implementing reinforcement learning (RL) in 3D, treating building volumes as intelligent agents and using points in geometry to manipulate their movements in an urban environment driven by sustainability criteria. Buildings interact with the environment of our cities giving us a diverse range of possible solutions to optomise the architecture of a building.

Additionally, we will be investigating methods to reduce the computational calculation time required for sunlight analysis; primarily by creating a reward-based system using RL. The application was created to reduce computational time and power requirements. It also resulted in a solution that increased the optimisation of the building volumes by 142% as based on sunlight analysis.