Automated generation of user activity-space pairs in space-use analysis

Automation of space-use analysis consists of two parts: automated generation of activity-space pairs and automated computation of space utilization based on the pairs. However, existing user presence models require architects to manually generate activity-space pairs because space-use types and their impact on the generation have not been formalized in a single method. This paper presents a formalized method that enables a computer to map activities onto spaces to generate such pairs by choosing spatial requirements of an activity, finding appropriate spaces, and mapping the activity onto found spaces. Validation tests show that activity-space pairs generated by this method conform to the pairs generated by expert architects, and the method is acceptable from the architects' perspectives. Automated space-use analysis powered by this method significantly enhances speed and consistency of utilization prediction. This paper contributes to virtual design and construction by providing a means to connect activities with spaces automatically.