Production System Loading–Cycle Time Relationship in Residential Construction

Abstract

Production home building possesses characteristics similar to manufacturing processes, such as the construction of more or less similar houses repeatedly and a growing demand for mass customization of homes. As a result of these similarities, larger homebuilders often attempt to view their production system as an assembly line process. However, the management tools generally utilized by these home builders are those used in other sectors of the construction industry, such as critical path method scheduling, cost estimating, and earned value analysis. These management tools do not provide an explanation or control/prediction tools for many undesirable situations that arise during home building, such as increasing cycle time which slows delivery of product to consumers and increases project capital costs, and increasing amounts of work in process that increases capital investment and thereby decreases company financial performance. In order to bring better management tools to the residential construction industry, this study examines relationships between cycle time, work in process, system throughput, new construction starts, and the capacity of the production system using building permit data for new single family homes in Chandler, Ariz. The applicability of Little’s law, a basic equation used in factory production management models, to a residential production system is examined. This study shows a definite, predictable relationship between cycle time, work in process, and production system throughput. It provides a pathway for further study of production system characteristics that have historically not been included in construction management models, with the expectation of developing new construction management tools that will account for more of the characteristics of construction production systems that affect project performance and company financial performance.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 48 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.