https://doi.org/10.24928/2026/0311
This research introduces a novel method for predicting project completion in construction by simulating and analyzing historical progress performance data through Markov Chain Analysis (MCA). This approach diverges from conventional regression analysis and quantitative risk analysis (QRA) via Monte Carlo simulation, which primarily emphasize modeling activity durations. Instead, the study evaluates progress by focusing on deliverables rather than activities, defining completion states instead of durations. The mutually exclusive states utilized are not-started (NS), work-in-process (WIP), and complete (OK). The method enables the application of MCA to determine when all deliverables achieve the absorbing state of completion, integrating confidence intervals and data discrimination criteria, while exhibiting satisfactory consistency in forecasting completion ranges.
WIP, completion, Markov chain analysis, machine learning.
Download: BibTeX | RIS Format
Reference in APA 7th edition format:
Samaniego, O. A.. (2026). Completion assessment for multiple production system using Markov chain analysis. In Hamzeh, F., Poshdar, M., & Garcia-Lopez,, N. P. (Eds.), Proceedings of the 34th Annual Conference of the International Group for Lean Construction (IGLC 34) (pp. 1737–1748). https://doi.org/10.24928/2026/0311
Shortened reference for use in IGLC papers:
Samaniego, O. A.. (2026). Completion assessment for multiple production system using Markov chain analysis. IGLC34. https://doi.org/10.24928/2026/0311