Despite substantial efforts to improve construction safety training, the accident rate of migrant workers is still high. One of the primary factors contributing to the inefficacy of training includes information delivery gaps during training sessions (knowledge-transfer). In addition, there is insufficient evidence that these training programmes alone are effective enough to enable migrant workers to transfer their skills to the jobsite (training-transfer). This research attempts to identify and evaluate additional interventions to improve the transfer of acquired knowledge to the workplace. For this purpose, this study presents the first known experimental effort to assess the effect of interventions on migrant work groups in a multinational construction project in Qatar. Data analysis reveals that the adoption of training programmes with the inclusion of interventions significantly improves training-transfer. Construction safety experts can leverage the findings of this study to enhance training-transfer by increasing workers' safety performance and hazard identification ability.

Construction safety requires real-time monitoring due to its hazardous nature. Existing vision-based monitoring systems classify each frame to identify safe or unsafe scenes, often triggering false alarms due to object misdetection or false detection, which reduces the overall monitoring system’s performance. To overcome this problem, this research introduces a safety monitoring system that leverages a novel temporal-analysis-based algorithm to reduce false alarms. The proposed system comprises three main modules: object detection, rule compliance, and temporal analysis. The system employs a coordination correlation technique to verify personal protective equipment (PPE), even with partially visible workers, overcoming a common monitoring challenge on job sites. The temporal-analysis module is the key component that evaluates multiple frames within a time window, triggering alarms when the hazard threshold is exceeded, thus reducing false alarms. The experimental results demonstrate 95% accuracy and an F1-score in scene classification, with a notable 2.03% average decrease in false alarms during real-time monitoring across five test videos. This study advances knowledge in safety monitoring by introducing and validating a temporal-analysis-based algorithm. This approach not only improves the reliability of safety-rule-compliance checks but also addresses challenges of misdetection and false alarms, thereby enhancing safety management protocols in hazardous environments.

The construction sector has long been known for its complicated safety requirements, which are critical to ensure the well-being of workers on site. However, interpreting these requirements and maintaining compliance can be difficult due to the amount and complexity of the paperwork involved. This leads to difficulty extracting safety information from requirement documents. Besides, information that is presented in a manner similar to human-like responses can improve employee understanding. This study proposed a Construction Safety Query Assistant (CSQA) approach to enhance the extraction and knowledge of construction safety regulations using Large Language Models (LLMs). CSQA comprises of three primary components: (1) the Construction Safety Investigation Module (CSI), which gathers and processes safety regulation documents through text extraction and preprocessing to build a searchable database; (2) the Safety Condition Identification Module (SCI), which utilizes LLMs to interpret user queries and extract relevant information from the CSI database, capitalizing on the models' ability to understand context and subtle textual nuances; and (3) the Safety Information Delivery Module (SID), which presents the retrieved information to users and integrates a feedback loop to refine the accuracy and relevance of responses based on user interaction. The CSQA approach was validated with 2 case studies that offered more contextually relevant, possibly lowering non-compliance risks, improving worker safety, and simplifying the consultation process in the construction sector. This study emphasizes its potential to transform access to crucial safety information in the construction industry.

The provision of suitable training materials is crucial in preparing construction workers for the safety risks inherent in their work environments. However, retrieving relevant training resources for specific work tasks and worksites poses challenges due to the dispersed, unstructured, and multifarious nature of available content. This paper proposes a linked data and ontology-based system framework for effectively sharing and accessing educational safety contents from heterogeneous sources in a standardized way. The main technical solutions in the framework include an ontology that leverages the Resource Description Framework and Ontology Web Language to structure and publish educational safety training contents based on the linked data principles, and a content retrieval method which leverages the SPARQL protocol and RDF query language. Experimental results confirm that the proposed approach outperforms existing repositories in terms of search time, search result precision, and accuracy. It is envisaged that the framework would aid safety trainers in sharing, searching, and retrieving didactically appropriate materials to prepare workers for the unique safety challenges of their jobsite environments.

Mobile scaffolding is essential in construction but presents significant safety risks, particularly falls from height (FFH) due to improper use and insufficient monitoring. While prior research has identified hazards, it often lacks robust, actionable solutions, especially regarding the comprehensive analysis of worker behaviors and the spatial context. This study proposed a computer vision-based safety monitoring system that leverages depth cameras for accurate spatial assessments and incorporates temporal conditions to reduce false alarms. The proposed system extends object detection algorithms with mathematical logic derived from safety rules to classify four key unsafe conditions related to safety helmet use, guardrail and outrigger presence, and worker overcrowding on mobile scaffolds. A diverse dataset from multiple sources enhances the model’s applicability to real-world scenarios, while a status trigger module verifies worker behavior over a 3 s window, minimizing detection errors. The experimental results demonstrate high precision (0.95), recall (0.97), F1-score (0.96), and accuracy (0.95) for safe behaviors, with similarly strong metrics for unsafe behaviors. The qualitative analysis further confirms substantial improvements in worker position detection and safety compliance using 3D data over 2D approaches. These findings highlight the effectiveness of the proposed system in improving mobile scaffolding safety, addressing critical research gaps, and advancing construction industry safety standards.