Construction is an important sector of every economy. Evidence of below par performance in construction projects has been recognized by government and industry bodies. Traditional control systems with project-based approaches have not overcome endemic problems in the industry such as cost and schedule overruns and quality issues. The innovative control system proposed in this research takes a production-based approach (as opposed to a project-based approach). The FULFIL system, aims to stabilise the work f low, minimise interruptions caused by q u a l ity problems and maximise the f lexib il ity in process design. The FULFIL system of production control is based on four pillars: queuing theory, transformation/flow/value theory, factory physics, and theory of constraints. In order to propose the principles of the FULFIL system, analytical and simulation models of construction production are developed. This thesis is driven by seven research objectives: 1) To analyse the impact of workflow variability on construction production. The research results confirm that performance in construction is adversely affected by workflow variability caused by factors such as rework and capacity imbalance. 2) To establish a tailored modelling approach that precisely quantifies variability in the workflow amongst specialty contractors. This thesis proposes a new modelling approach using a relative indicator of variability, which takes both the standard deviation of time between completions and average processing times into consideration. 3) To explore approaches to stabilising the workflow in construction. Two principles for stabilising the workflow are proposed and tested. Limiting the number of jobs under construction and integrating work processes are confirmed to prevent frequent work starvations and overloads in the production network. 4) To explore opportunities for variability reduction in construction. Tangible performance measures in due-date-driven and rate-driven production are compared. FULFIL analysis shows that when new construction is authorised, not scheduled, the production is more efficient, controllable and robust against control errors. 5) To explore opportunities for variability buffering in construction production. The user-friendly framework for defining optimum-sized capacity buffers in the FULFIL system is developed and tested. 6) To explore opportunities for improving the flexibility in construction processes. Two sources of inflexibility in process designs are analysed and addressed. Depending on the level of capacity imbalance and processing time variability, different cross-training strategies are proposed and tested. When processing times are variable, capacity should be shifted in an indirect path to the bottlenecks. 7) To explore opportunities for reducing interruptions caused by quality problems and rework. Three variables of rework are analysed and strategies to address them are proposed. Rework duration and intervals, and the timeframe of call-backs are shown to have significant impacts on the performance of construction and can be effectively offset by using FULFIL protocols. This thesis contributes to the body of knowledge by developing a deeper insight into the dynamics of workflow, quality and flexibility management, and the resulting impacts on construction plan reliability. Furthermore it can assist industry practitioners in finding the most cost-effective way to operate and control production networks. Easy-to-use models developed and tested in this thesis can improve the traditional project-based controls in construction.
展开▼
