Scenario for embedding Al in Acoustic Design. Exploiting applications at several design stages

摘要

Holistic methods for acoustic design project are increasingly developed, combining the knowledge gained from physical acoustic researches, early stage noise mapping, architectural parameters and psychoacoustic. The aim of these methods is to combine subjective and objective parameters to achieve more performative outcomes to satisfy comfort needs. These methods require a methodological instrument to structure the design problem and to develop a heuristic mechanism acquiring broad rules to handle the specific solution. Today, digital tools handle large complex problems evolving from generative into intuitive tools such as the Artificial Intelligence (AI). AI, through the deep learning, discovers intricate structure in large data sets, by using the backpropagation algorithm, to indicate machine's way to change its internal parameters thus helping to develop unpredictable innovative solutions. The paper intends to develop a theoretical background based on bibliographic survey of AI in architecture, with a focus on acoustics. The resultant reading grid will be implemented by a custom survey sent to main stakeholders of acoustic design aiming at understanding strength and weaknesses of its application.