ABSTRACT
The way in which buildings are designed plays a critical role in responding to the climate emergency. It is estimated that Architecture, Engineering and Construction industry is responsible for up to 38% of the Global Greenhouse Gas emissions. During the early design phases, the possibility to adjust the design is usually high, but the data concerning the project is scarce. Architects need tools to help them make informed decisions early in the design process, and artificial intelligence may be useful in that respect. A trained Neural Network has been developed to predict multi-family building carbon footprint based on previously simulated training data. The article describes the process of creating the model, preparing the training data, training the neural network and using it within the Grasshopper environment.
