Simulation-based and statistical analysis of the learning effect in floating caisson construction operations

Antonios Panas, John Paris Pantouvakis

Research output: Contribution to journalArticlepeer-review

11 Citations (Scopus)


The learning curve theory is implemented for the assessment and prediction of performance in repetitive and complex construction operations. The purpose of this study is to explore the applicability of learning curve concepts for the fabrication of floating caissons, a construction activity that has not been adequately explored in the published literature. Although standard statistics and simulation methods have been used, a direct comparison of the produced results, along with their emerging inferences, is attempted for the first time. In addition, simulation modeling is used to predict future performance within the learning curve paradigm. An extended version of an existing simulation platform (CaissonSim) is used to execute the computations. A straight-line model is applied to quantify the improvements in productivity and the dynamically changing learning rates. A comparative analysis of the results has shown that both the statistical and simulation-based approaches have yielded satisfactory results, although the level of accuracy and variability is differentiated. Thus, it is suggested that the synergetic implementation of both methods of analysis is more beneficial because, when combined, they provide a robust analytical framework.

Original languageEnglish
Article number04013033
JournalJournal of Construction Engineering and Management
Issue number1
Publication statusPublished - Jan 1 2014


  • Caissons
  • Productivity
  • Quantitative methods
  • Simulation
  • Statistics

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Industrial relations
  • Strategy and Management

Fingerprint Dive into the research topics of 'Simulation-based and statistical analysis of the learning effect in floating caisson construction operations'. Together they form a unique fingerprint.

Cite this