Predicting Cost and Duration in Construction Projects using Artificial Intelligence
Article Sidebar
Main Article Content
Abstract
Accurately estimating the cost and duration of construction projects remains a recurring challenge for effictive project management. This study proposes an Artificial Intelligence (AI) and Machine Learning (ML)-based approach to improve the prediction of these critical parameters, comparing its performance to traditional methods such as CPM, PERT, and Monte Carlo simulations. The research incorporates linear regression models, Random Forest, XGBoost, and automated algorithms (AutoML) applied to both synthetic and real datasets. The results show that ML-based models, particularly Extra Trees and XGBoost, achieve coefficients of determination above 0.92 for cost prediction, significantly surpassing conventional techniques. However, duration prediction is influenced by the quality of the available data, highlighting the need for more complete and representative datasets. It is concluded that AI enables the optimization of planning and decision-making in construction projects, and that the creation of a standardized, centralized database—integrated into the IDE Misiones platform—would be strategic in strengthening the sector’s analytical and predictive capabilities in the province.
Downloads
Article Details
Section
References
Djatnika, Suntana Sukma, Witjaksana, Budi, Oetomo, Wateno, y Bambang, W., «Risk Analysis on the Construction Project of the Basement Building under Jalan Pemuda-Yos Sudarso, Surabaya», J. Phys. Conf. Ser., vol. 1364, 2019, doi: 10.1088/1742-6596/1364/1/012073.
Ullah, K, Abdullah, A. H., Nagapan, S., Suhoo, S., y Khan, M. S., «Theoretical framework of the causes of construction time and cost overruns», IOP Conf. Ser. Mater. Sci. Eng., vol. 271, 2017, doi: 10.1088/1757-899X/271/1/012032.
Anastasia Stepanets, «Método de la ruta crítica en la gestión de proyectos», https://blog.ganttpro.com/. [En línea]. Disponible en: https://blog.ganttpro.com/es/metodo-de-la-ruta-critica-en-la-administracion-de-proyectos/. Accedido el: 28/10/2025.
Team Asana, «El diagrama de PERT: qué es y cómo crearlo (incluye ejemplos)». [En línea]. Disponible en: https://asana.com/. Accedido el: 28/10/2025.
Sonia Coma, «Método de simulación Montecarlo: Qué es y cómo adoptarlo en tu empresa». [En línea]. Disponible en: https://asana.com/. Accedido el: 28/10/2025.
Setiawan, A, Fadjar, A., y Labombang, M., «Scheduling the Construction of Low-Income Community Houses in Palu City Using the Probabilistic Duration Method», IOP Conf. Ser. Earth Environ. Sci., vol. 1355, 2024, doi: 10.1088/1755-1315/1355/1/012013.
Mirindi, Derrick, Mirindi, Frederic, Bezabih, Tajebe, Sinkhonde, David, y Kiarie, Winfred, «Review: The Role of Artificial Intelligence in Building Information Modeling», ACM- SIMIS-CPR, 2025, doi: 10.1145/3716489.3728433.
Nihad Hassan, «AI vs. machine learning vs. deep learning: Key differences», https://www.techtarget.com/. [En línea]. Disponible en: https://www.techtarget.com/searchenterpriseai/tip/AI-vs-machine-learning-vs-deep-learning-Key-differences/?utm_source=chatgpt.com. Accedido el: 28/10/2025.
I. H. Sarker, «Deep Learning: A Comprehensive Overview on Techniques, Taxonomy, Applications and Research Directions», SN Comput. Sci., vol. 2, 2021, doi: 10.1007/s42979-021-00815-1.
Demiss, Belachew A. y Elsaigh, Walied A., «Application of novel hybrid deep learning architectures combining Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN): construction duration estimates prediction considering preconstruction uncertainties», Eng. Res. Express, vol. 6, 2024, doi: 10.1088/2631-8695/ad6ca7.
Soreti M Liben, Demiss A Belachew, y Walied A Elsaigh, «Comparing advanced and traditional machine learning algorithms for construction duration prediction: a case study of Addis Ababa’s public sector», Eng. Res. Express, vol. 6, 2024, doi: 10.1088/2631-8695/ad979f.
Mohammad Savargiv, Behrooz Masoumi, y Mohammad Reza Keyvanpour, «A New Random Forest Algorithm Based on Learning», Comput. Intell. Neurosci, 2021, doi: 10.1155/2021/5572781.
Zi-Yi Yang, Zhao-Feng Ye, Chang-Yu Hsieh, y Yi-Jia Xiao, «SPLDE XTRA T REES : R OBUST MACHINE LEARNING APPROACH FOR PREDICTING KINASE INHIBITOR RESISTANCE», arXiv preprint, 2021.
Cynthia Lokker, Wael Abdelkader, y Elham Bagheri, «Boosting efficiency in a clinical literature surveillance system with LightGBM», PLOS Digit. Health, 2024, doi: /10.1371/%20journal.pdig.0000299.
Chi Wang, Qingyun Wu, Markus Weimer, y Erkang Zhu, «FLAML: A F AST AND L IGHTWEIGHT AUTO ML LIBRARY», Proc. 4 Th MLSys Conf., 2021.
Arunn Thevapalan, «The Machine Learning Workflow Explained (and How You Can Practice It Now)». [En línea]. Disponible en: https://towardsdatascience.com/. Accedido el: 28/10/2025.
Deepakcode07, «Construction-Project-Data-Generator». https://github.com/. [En línea]. Disponible en: https://github.com/Deepakcode07/Construction-Project-Data-Generator?utm_source=chatgpt.com. Accedido el: 28/10/2025.
«JohnVans123/ProjectManagement». [En línea]. Disponible en: https://huggingface.co/datasets/JohnVans123/ProjectManagement?utm_source=chatgpt.com. Accedido el: 28/10/2025.
Mohammad Rafiei, «Residential Building». [En línea]. Disponible en: https://archive.ics.uci.edu/dataset/437/residential%2Bbuilding%2Bdata%2Bset?utm_source=chatgpt.com. Accedido el: 28/10/2025.
Google, NotebookLM. [En línea]. Disponible en: https://notebooklm.google/. Accedido el: 28/10/2025.