Discipline-Sensitive Predictive Analytics for IPA-Driven Building Maintenance Management: Material Stock Quantity Modeling

Abstract

The study introduces a machine learning model for BMM (Building Maintenance Management), which utilized IPA (Intelligent Process Automation) to predict the material stock required in a period, to manage the cost efficiency. Traditional BMM approaches often prepared not efficient amount of material stock for pending maintenances required materials, that could be resolved by this machine learning model; it uses discipline-sensitive machine learning algorithms to predict material needs accurately and ensure the efficient amount of material stock levels and reducing the financial waste.

A proof-of-concept is implemented in a case study from this industry that validates the machine learning model's effectiveness, showing the capability of combining Artificial Intelligence (AI) and machine learning to apply the material stock level prediction. The proposed approach not only predicts the material stock efficiency but also ensures the methodology of AI that could be utilized in BMM, which could be further attached to smart urban development. With the support of the case study, the introduction of this approach could make the BMM field a significant advancement in AI-powered material stock management. It resolved both manufacturer production waste and unnecessary finance cost, helping reduce resource waste and promoting sustainability.

The outcomes of this research include improved accuracy in forecasting material stock levels for upcoming building maintenance tasks, explore in this area with more sophisticated AI algorithms, mathematic, and analytics, also considered the potential IoT integration in the material warehouse to analyze and predict in real time material stock level. In a broader context, we are introducing a new standard of material stock management that could collaborate with AI together to enhance the BMM in a smart urban environment.

Received: 25 July 2024 | Revised: 31 March 2025 | Accepted: 29 April 2025 

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study. Building maintenance companies could request the author for the potential implementation based on their own confidential data.

Author Contribution Statement

Zhimeng Huang: Conceptualization, Methodology, Software, Validation, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Project administration. Xiaodong Liu: Conceptualization, Methodology, Validation, Investigation, Resources, Writing – review & editing, Supervision, Project administration. Imed Romdhani: Conceptualization, Conceptualization, Supervision, Project administration.