On the Origin of Occupant Behavior: The Evolution of Simulating Effects of People in the Building Performance

Abstract

The use of computers in the design of buildings predates the invention of the electronic computer with civil engineer Konrad Zuse designing the Z3 electromechanical computer for a number of uses including structural engineering calculations. As all electronic computers replaced electromechanicals, the use of computers in building simulation grew. Starting with BRIS in Sweden in 1963 and the “Post Office Program” in the U.S. in 1971, building performance simulation, the simulation of thermodynamics, airflow, lighting, and acoustics, has grown enormously, with hundreds of programs now available for designers ranging from simple spreadsheets to software that can do full 3D heat transfer and computational fluid dynamics simulation. While the thermodynamics and airflow are now simulated with surprisingly high fidelity, the effects of people are usually not. Whereas people are diverse, unpredictable, and reactive, the actions of people are usually simulated as homogenous, predictable, and non-reactive. This talk will focus on the evolution of including the effects of people on building performance and how the use of big data, agent-based modeling and artificial intelligence (AI) might finally start to get it right.