抄録
This study examines the previously unknown relationship between using human behavior simulation, equipped autonomous, intelligent Virtual-Users, and students′ self-experimentation performance in fire egress planning. The research method involved 70 students in authentic design courses who proposed floor plans for office buildings before and after simulating evacuee behaviors. They then scored their experiences based on using the simulation. Statistical analysis of those scores reveals that using human behavior simulation helps students find unexpected problems, evaluate the validity and functionality of design solutions, conduct the experimentation process more efficiently, and determine the solutions with relative ease. The main reasons for these results are posited to be the explicit, analytic, and observable representation of virtual evacuees, their manipulative parameters, and an integrated system between human behavior simulation and Building Information Modeling (BIM). The findings of the present study can contribute to developing a rational computational means for education in fire egress planning.
