|Ph.D Thesis||Department of Civil and Environmental Engineering|
|Supervisors:||Assoc. Prof. Sacks Rafael|
|Assoc. Prof. Shapira Aviad|
|Full Thesis text|
Interior and finishing activities in building construction exhibit high degrees of variation as a result of uncertainty in supply chains, variations in work quantities, client changes and lack of predictability of the production capacity of subcontracting trades. Decisions must constantly be made concerning effective utilization of available resources. Re-entrant workflow patterns, where a trade crew returns twice or more to the same space, make production control particularly difficult.
Modeling and managing the flow of interior finishing works in building construction has been hampered by the inability of traditional critical path modeling to reflect features of these works such as uncertainty, instability, non-linear and interrupted value-adding processes and re-entrant flow. Drawing on empirical data collected in a work study of a large residential project, a novel workflow model is proposed that explicitly models the flows of products and crews at a fine-grained level. The overall model includes a Trade Workflow Model (TWM) with a module for each work-package type, and a Project Workflow Model (PWM) in which multiple trade modules are embedded. The model was tested through implementation of a simulation of a prototypical construction project involving seven work-packages performed in a 20-storey residential tower with 120 customized apartments. The modeled flows of crews through the test building reflected the same behavioral features observed in construction projects. The model enables evaluation of the impact of management policies on production flow at different levels of detail. It is useful as a research tool and for future development of construction management software.
A method was proposed for implementation of lean principles of production flow shielding and pull flow control at the operational level, even where re-entrant flow occurs. Pull of works is achieved through real-time prioritization of pending work-packages and daily regulation of crew assignments and trades’ production capacities. Application of various heuristics was evaluated using discrete-event simulation of a representative construction project based on the Trade Workflow Model (TWM) and the Project Workflow Model (PWM). The experimental results emphasize the importance of dynamic control of allocation of production resources to those mature activities that ensure stable upstream and downstream production flow. This is particularly the case with re-entrant flow patterns. The most successful policy was to balance the trade with re-entrant flows to ensure sufficient feeding of new work to successive crews.