TR185: An Automated Method to Identify Occupant Interactions in Renovations of Occupied Buildings

Renovation projects represent an increasing percentage of building construction projects today. As the importance to utilize existing resources responsibly becomes greater, building owners are finding ways to extend the life of existing buildings (e.g., through systems upgrades), while still maintaining tenant business operations. Renovations of occupied buildings create particular management challenges because tenants and construction crews must share spaces in the building. These occupant interactions can at best be tolerable, and at worst, be disruptive to tenants and crews. Todayís practice of manually identifying occupant interactions misses many disruptive interactions, which can lead to cost and schedule overruns and business disruptions for the building occupants. Therefore, this thesis addresses this challenge by developing an automated method to identify occupant interactions (IOI method).
Identifying occupant interactions is challenging because it requires the integration of spatial, organizational, and temporal information to understand where, when, and how occupants utilize building spaces over different workshifts during renovation. Furthermore, the dynamic nature of renovation projects causes the building configuration (i.e., locations of occupants) to change many times during renovation. Therefore, this thesis advances prior work in product, organization, and process modeling by formalizing the relationships between spatial, organizational, and temporal renovation planning information to enable a thorough (i.e., analysis of all building configurations) and detailed (i.e., at the space and workshift level) identification of occupant interactions. The IOI method formalizes occupants with specific organizational requirements, processes at the level of detail required for accurate identification of occupant interactions, process-space relationships to simplify model-based representation of tenant move activities, and organizational relationships that represent how occupants share spaces. These formalizations enable reasoning methods that update occupant locations and their space sharing abilities and identify interactions at a level of detail that is infeasible to achieve with manual methods. The IOI method is also representationally more efficient than existing computer-based methods to identify occupant interactions.

I implemented the IOI method in a computer prototype, 4DRenCheck, to test that the method supports automated and accurate identification of occupant interactions. I then validated the method prospectively with three renovations of large office buildings that were in their planning stages. Prospective validation tested the power of the IOI method to identify interactions more accurately, thoroughly, and in greater detail than traditional methods and tested whether the method could be implemented in a timeframe to affect future project decisions. Based on the insights from validating the IOI method, this thesis also provides a six-criterion framework to determine when prospective validation is an appropriate method for researchers to use. The framework relates the objectives of virtual design and construction (VDC) methods (i.e., predict project performance, use on design-construction projects, and support business objectives) to six validation parameters that can demonstrate this power. This thesis also provides researchers with implementation guidelines on planning, executing, and analyzing a prospective validation test.

The results of the validation demonstrate that the IOI method analyzed all unique building configurations at the space and workshift level of detail. Consequently, the method identified occupant interactions more accurately than with traditional methods. Based on the insights from 4DRenCheck, the following interventions were made: one renovation planner corrected the start and end locations of tenant move activities, another renovation planner intended to revise the construction schedule to a greater level of detail, and the third renovation planner anticipated revising the sequencing of the renovation activities to better utilize space in the building. The results demonstrate that the concepts formalized in this thesis provide the power necessary for thorough, detailed, and accurate identification of occupant interactions on full-scale renovation projects. More broadly, the IOI method demonstrates that the integration of product, organization, and process models and the automation of planning tasks, such as identifying occupant interactions, is beneficial in managing towards desired performance objectives in renovation projects.