This chapter introduces a Sustainable Project Management in
Construction framework that operationalises artificial intelligence to (i)
anticipate schedule slippages, (ii) optimise labour and material deployment,
and (iii) embed sustainability indicators directly into project
decision-making. Using records from 33 projects spanning infrastructure,
residential, and commercial sectors, we develop and validate a Random Forest
Regressor benchmarked against alternative machine-learning models to forecast
schedule overruns and resource utilisation. A 10-fold cross-validation
protocol, residual diagnostics, and scenario analyses demonstrate robust
predictive performance for the delay prediction: R²≈0.87, MSE≈11.8 days;
resource utilisation: R²≈0.82 without degrading when sustainability features
are included. We describe a practitioner-oriented, Python-based interface that
ingests site data, automates preprocessing, and returns interpretable forecasts
alongside waste-reduction and recycled-content metrics to support “what-if”
analyses. Governance measures anonymisation, role-based access, and compliance
with institutional protocols, are integrated to protect stakeholders and
sustain data trust. The chapter situates SPMC within current literature and
practice, distilling design choices and implementation patterns that readers
can adapt to heterogeneous project contexts. Limitations and future directions
are discussed, including expanding the dataset’s diversity, coupling with IoT
data streams for real-time sensing, and integrating life-cycle assessment to
widen coverage of carbon and circularity indicators. Collectively, SPMC
provides a replicable pathway for aligning cost-time performance with
measurable environmental outcomes in contemporary construction management.
Author(s) Details
Mohamed Y. Laissy
Department of Civil Engineering, University of Prince Mugrin, Medina, Saudi
Arabia.
Omar Mostafa Dakhil
Department of Architecture Engineering, University of Prince Mugrin,
Medina, Saudi Arabia.
Please see the book here :- https://doi.org/10.9734/bpi/erpra/v10/6163
No comments:
Post a Comment