Automated hazardous area identification using laborers' actual and optimal routes
Introduction
The construction industry faces more risk factors than any other industry because each construction site has its own individual characteristics [1], [2]. Compounding such complexity is the fact that construction accidents can be divided into two categories: those occurring while working and those occurring while moving to the working space. The fatal accident reports of the Korea Occupational Safety and Health Agency (KOSHA) [3] indicate that approximately 20% of fatal accidents in the construction industry occur when laborers are moving through a site. According to the Health and Safety Executive [4], 23% of all accidents in the construction industry occur while the worker is moving through a site. The risks that laborers face in their movement path are significantly different from the risks that laborers face while in their working places; this is due to the performance of other tasks, the piling of risky materials, the existence of openings in the floor, etc. Furthermore, since a construction schedule is established based on activities, safety management focuses on work spaces [5], [6], which in turn means that the level of management for moving processes or paths is usually lower than for work spaces. The dynamic changes within construction sites also make it hard for safety managers to identify the beginning and the end of the overlap of hazards and the laborers'movement paths [7].
Carter and Smith [8] suggested that the hazard identification ratio of general projects is 66.5%. This means that the unidentified hazard ratio is greater than 30%. To detect unidentified hazards, additional resources (e.g., increasing the number of safety managers) are required to improve hazard identification under this current practice. However, even though unidentified hazards entail more risk than identified hazards, it is not economical to input additional efforts (i.e., increasing the number of safety managers) to identify more hazards. Additional resources may not be available for several reasons, such as financial limitations, lack of experts, etc. Regardless of why a site cannot add additional resources toward hazard identification, leaving risks unidentified means that the exposure time for a hazard can increase. For this reason, a method that can decrease the exposure time of hazards is required.
An automated approach—such as information technology (IT)—offers an alternative for identifying hazards promptly. A large amount of research deals with the IT-based location-tracking approaches that improve safety [9], [10], [11]. Hallowell et al. [12] used proximity sensors—consisting of radio frequency identification (RFID) and ultra-wideband sensors—to prevent struck-by-vehicle accidents. Navon and Kolton [9], [11] suggested a fall-prevention model based on the location tracking of safety equipment. In addition, Lee et al. [13] developed a laborer location-tracking-based system that warns laborers when they approach hazardous areas. Although these approaches offer value in risk mitigation, since they all focused on risk control and assumed that the hazards were already identified, they are not likely to accurately manage risks and hazards when such dangers are unidentified.
Therefore, the objective of this study is to develop a system that automatically identifies hazards in laborers' movement paths through laborer location tracking. The system identifies hazardous areas using the deviation between laborers' location logs and laborers' paths. By comparing existing hazardous areas and work spaces, previously unidentified hazardous areas can be identified. The proposed system uses real-time tracking to identify hazards that were previously unknown and to reduce the duration that such hazards remain without any safety countermeasures.
The rest of this study is organized as follows. First, this study investigates traditional hazard identification approaches. Then the related research for hazard identification is presented, and various methods are analyzed. From this analysis, assumptions for developing algorithms and a conceptual model are suggested. Finally, the detailed algorithms of the suggested system are explained with a validation of the system.
Section snippets
Literature review
To prevent accidents, hazard identification is necessary [3]. Unidentified hazards mean that hazards are not included in the safety management process. The traditional way of identifying construction hazards is through a rule-based checklist established based on accident cases and best practices. Traditional hazard identification methods have failed to identify all of the hazards that should have been identified [3] because hazards are generated by a combination of unexpected conditions in
Hazard identification model
This section presents the approach we used to organize the modules and related databases (DBs) to establish a hazard identification model. We begin by discussing the assumptions used when developing the model, then we cover the algorithms we developed for each module.
System development
The purpose of this section is to develop a system by using a library developer kit, an API developer kit, and an RTLS device. Before system development, the protocol—the form of data used to transmit data between each RTLS device—should be defined. The detailed contents of these protocols and devices are shown in Table 1, Table 2.
A two-step approach is taken to develop the system. First, BIM property values are extracted using a developer's kit. In this study, ArchiCAD 13—an easy tool to use
Case study results
Accuracy and promptness can be considered two of the most important criteria for a safety management system. Promptly identifying hazardous areas on laborer's paths empowers safety managers to respond to these areas. It can also reduce the time in which a hazardous area is left without any countermeasure, which decreases the degree of laborers' exposure to a hazard. A highly accurate hazard identification system provides the precise coordinates of a hazard. By providing information about a
Conclusions
Currently, safety management is executed to focus on the work space. Because of the inherent risks in the construction industry, there are hazards on construction sites not only in work spaces but also in pathways. Moreover, unidentified hazards increase the probability of an accident. Although safety managers perform hazard identification to prevent accidents, it is difficult for them to identify every single hazard.
This study develops a model that determines hazardous areas on a laborers'
Acknowledgments
This research was supported by a grant from the BIM R&D Program (14AUDP-C067809-2) funded by the Ministry of Land, Infrastructure, and Transport of the Korean government.
References (30)
- et al.
Visualizing safety assessment by integrating the use of game technology
Autom. Constr.
(2012) - et al.
Building information modeling (BIM) and safety: automatic safety checking of construction models and schedules
Autom. Constr.
(2013) - et al.
Situational awareness of construction equipment using GPS, wireless and web technologies
Autom Constr.
(2003) - et al.
Revealing the “Invisible Gorilla” in construction: Estimating construction safety through mental workload assessment
Autom. Constr.
(2016) The relationship between employees' perceptions of safety and organizational culture
J. Saf. Res.
(2002)- et al.
Methodology for integrated risk management and proactive scheduling of construction projects
J. Constr. Eng. Manag.
(2008) - et al.
Activity-based safety risk quantification for concrete formwork construction
J. Constr. Eng. Manag.
(2009) Annual Report: Fatal Accident Cases
(2001–2003)- et al.
Developing an Effective Safety Culture: A Leadership Approach
(2002)
Scheduling-based risk estimation and safety planning for construction projects
J. Constr. Eng. Manag.
Spatial and temporal exposure to safety hazards in construction
J. Constr. Eng. Manag.
Safety hazard identification on construction projects
J. Constr. Eng. Manag.
Model for automated monitoring of fall hazards in building construction
J. Constr. Eng. Manag.
Simulation of automated data collection in buildings
J. Comput. Civ. Eng.
Cited by (70)
Real-time risk assessment of multi-parameter induced fall accidents at construction sites
2024, Automation in ConstructionDigital technologies in construction: A systematic mapping review of evidence for improved occupational health and safety
2023, Journal of Building EngineeringSpatial-temporal analysis of safety risks in trajectories of construction workers based on complex network theory
2023, Advanced Engineering InformaticsDynamic human systems risk prognosis and control of lifting operations during prefabricated building construction
2023, Developments in the Built EnvironmentAutomated detection of dangerous work zone for crawler crane guided by UAV images via Swin Transformer
2023, Automation in ConstructionSimulating travel paths of construction site workers via deep reinforcement learning considering their spatial cognition and wayfinding behavior
2023, Automation in ConstructionCitation Excerpt :In general, construction workers prefer to choose the shortest path for work efficiency and to minimize physical exertion. However, if the shortest paths are in an area where work occurs frequently or where hazardous objects are present, workers may choose a different path; this tendency is identified in the experiments of Kim et al. [60]. Also, according to the experiment of Yang et al. [61], the type of hazards can affect the movement path of workers.