Over the last decade, a considerable amount of research has documented the application of machine learning (ML) and its potential for cleaner production of sustainable construction materials particularly on geopolymers. Conceptually, the use of ML could help optimize the mixture composition, predict the property and performance of geopolymers materials. However, existing studies seem to mainly concentrate on geopolymer concrete and thus overlook other forms such as mortar and paste, and the data requirements of ML. In addressing the gaps, the aim of this study is to provide a current status of art on the use of ML on geopolymer materials by specifically exploring (i) the progression of ML in geopolymer materials from 2012-2023; (ii) the forms and types of geopolymer being researched using ML; (iii) the data sources and sizes, and ML algorithms being used; and (iv) the tasks being performed using ML. The overall findings show that ML are primarily utilized for predicting geopolymer properties, particularly compressive strength, while their potential in mixture optimization and structural maintenance remains largely untapped. Additionally, the small training datasets and the predominant reliance on data from previous publications in most studies underscore the limited utilization of field data. In conclusion, this study informs researchers of the current challenges in the application of ML for geopolymer materials and proposes directions for future research in using ML for improved property prediction and mixture optimization of sustainable geopolymer materials.

Ageing building stock, shrinking budgets, and inspector shortages hinder timely façade safety inspections. This research develops an automated damage detection and risk prioritization system for aging concrete structures. Five YOLOv11 variants were trained on 130,838 high-resolution images from 25 Seoul districts to detect three critical damage types: cracks, exposed rebar, and spalling. The proposed framework integrates YOLOv11 detection with a novel Damage Criticality Index (DCI) that transforms five visual-spatial cues—area, multiplicity, confidence, density, and spread—into continuous severity scores, subsequently categorized into low, medium, and high risk via K-means clustering. YOLOv11x achieved 0.78 mAP@0.5 at 101 FPS, enabling real-time processing suitable for field deployment. Field trials confirmed robust detection and consistent risk ranking in both uncluttered and cluttered urban environments, substantially reducing inspection time and minimizing missed defects compared to conventional manual methods. The framework provides scalable, data-driven support for city-wide monitoring and transparent, risk-prioritized maintenance of aging infrastructure.

Purpose The number of bidders in upcoming tenders has important managerial implications for both construction clients and contractors in their decision-making in the competitive bidding process. However, there is a stagnation of research efforts on predicting the number of bidders with only a handful of studies over the past decades, which mainly focused on statistical distribution of the number of bidders. This study aims to provide a new perspective of predicting the number of bidders using machine learning (ML) algorithms. Design/methodology/approach This study adopted a case study approach with a bidding dataset of public sector construction projects in Singapore. Six ML models were developed, and linear regression was used as a baseline model is assessing the predictive performance of ML models. Findings The results show that ML models outperform the baseline linear regression model, in which XGBoost is the best performing model of R2 which is two times higher than the linear regression model. In addition, economic-related factors play a vital role in this prediction problem. Research limitations/implications While the predictive performance of the developed ML models is relatively low, it indicates the challenges and complexities in this prediction problem, even with the use of artificial intelligent techniques. Originality/value Being a pioneering work, this study sets a foundation for the use of ML models in this prediction problem and offers insights for future modelling attempts towards the development of a decision support system for construction clients and contractors.

This research aims to identify and prioritize key planning elements for the redevelopment of such housing complexes by incorporating perspectives from both experts (supply-side) and residents (demand-side). To achieve this, a hybrid multi-criteria decision-making framework was developed by integrating the Analytic Hierarchy Process (AHP) and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). A total of 25 planning elements were identified through Focus Group Interviews and organized into five domains: legal and institutional reforms, project feasibility, residential conditions, social integration, and complex design. The AHP was used to assess the relative importance of each element based on responses from 30 experts and 130 residents. The analysis revealed a clear divergence in priorities: experts emphasized feasibility and regulatory considerations, while residents prioritized livability and spatial quality. Subsequently, the TOPSIS method was applied to evaluate four real-world redevelopment cases. From the supply-side perspective, Seoul A District received the highest score (0.58), whereas from the demand-side perspective, Gyeonggi D District ranked highest (0.69), illustrating the differing priorities of stakeholders. Overall, Gyeonggi D District emerged as the most favorable option in the combined evaluation. This research contributes a structured and inclusive decision-making framework for the regeneration of public housing. By explicitly comparing and quantifying the contrasting preferences of key stakeholders, it underscores the critical need to balance technical feasibility with resident-centered values in future redevelopment initiatives.