System and method for evaluating safety index of construction site worker

Abstract

A system and method for evaluating a worker's safety index through a model capable of quantitatively evaluating the risk or safety of a construction site worker are disclosed. According to one aspect of the present invention, a method for evaluating a safety index of a construction site worker is provided, comprising: a numerical model construction step of constructing a numerical model for evaluating the safety index of a construction site worker; an evaluation data acquisition step of the safety index evaluation system acquiring evaluation data for each of a plurality of influencing factors of the worker to be evaluated; a risk scale determination step of the safety index evaluation system determining a risk scale for each of the plurality of influencing factors of the worker to be evaluated based on the evaluation data for each of the plurality of influencing factors of the worker to be evaluated; and a safety index calculation step of the safety index evaluation system applying the risk scale for each of the plurality of influencing factors of the worker to be evaluated to the numerical model to calculate the safety index of the worker to be evaluated.

Description

Technology Field

[0001] The present invention relates to a system and method for evaluating a worker's safety index through a model capable of quantitatively evaluating the risk or safety level of a construction site worker.

[0002] This invention was produced through research conducted in 2022 with funding from the Ministry of Land, Infrastructure and Transport and support from the Korea Institute of Land, Infrastructure and Transport Technology (RS-2021-KA161440, Development of LOD350 digital twin application technology for disaster prevention equipped with AI object recognition, path calculation reflecting construction progress, and accident prediction algorithm for worker protection and facility safety management). Background Technology

[0003] The construction industry is a major sector that contributes to job creation and economic growth, but it is also a high-risk industry where accidents can result in significant casualties, including those involving outdoor, high-altitude, and heavy equipment operations. To ensure safety at construction sites, various laws, such as the Industrial Safety and Health Act and the Serious Accidents Punishment Act, are in effect. The Industrial Safety and Health Act supports the establishment of safe working environments by presenting detailed standards based on site conditions. Furthermore, to ensure strict compliance with safety standards at each site, the Serious Accidents Punishment Act was enacted, imposing severe penalties on business owners and management executives in the event of accidents at construction sites with a value exceeding 5 billion won or workplaces with 50 or more permanent employees (Ministry of Employment and Labor, 2022). Additionally, reflecting the multi-layered subcontracting structure of the construction industry, a Special Act on Construction Safety is currently under discussion. This act would impose safety responsibilities on all project stakeholders—including ordering, design, construction, and supervision—in the event of an accident.

[0004] Not only in Korea but also globally, there is a growing trend of providing policy support to actively utilize Fourth Industrial Revolution technologies in the construction industry to reduce safety accidents at construction sites; consequently, the market size of major smart construction technologies is gradually expanding. In Korea, various smart construction technologies, such as IoT, artificial intelligence, and big data, are being developed with the goal of reducing the number of fatalities in the construction industry by 30% by 2025 (Ministry of Land, Infrastructure and Transport, 2018). Following this trend, domestic construction companies are also implementing safety management through the application of various smart technologies on-site. Construction Company A has developed the Internet of Things (IBOT), which manages all risk factors in the basement via a Bluetooth network, to detect dangerous situations in real time and support the rapid evacuation of workers in the event of an emergency. Meanwhile, Construction Company B has developed an IoT-based Safety System to apply smart construction technologies to the site, aiming to prevent accidents and minimize fatalities in the event of an accident.

[0005] Despite various attempts to reduce safety accidents in the construction industry, the number of fatal accidents has not decreased significantly. Over the past five years from 2016 to 2020, an annual average of approximately 475 fatalities occurred in the construction industry, accounting for about 51.2% of the total industrial accidents. In 2021, the number of fatal industrial accidents was 828, with the construction sector accounting for 417 deaths—50.4% of the total—a proportion 28.2% higher than the manufacturing sector, which ranked second. The construction industry had the highest fatality rate per 10,000 workers (calculated by multiplying the number of deaths by 10,000 and dividing by the total number of workers) at 1.75, the highest among all industries and approximately 3.8 times higher than the manufacturing sector, which ranked second (Ministry of Employment and Labor, 2021). When comparing accidents based on project value, small-scale sites with a scale of less than 5 billion won accounted for 71.5% of the total; this indicates that more accidents continue to occur at small-scale sites that are not subject to the Serious Accidents Punishment Act. To assess the level of accidents in the domestic construction industry, a comparison was made with Japan, Germany, and the United Kingdom, which account for a similar proportion of the total industry while considering national income levels. The results showed that Korea exhibits a relatively high fatality rate per 10,000 workers.

[0006] Despite significant efforts to reduce safety accidents at construction sites, such as strengthening laws and regulations and introducing smart construction technologies, the construction industry’s accident rate remains high compared to other industries and developed nations, indicating the need for improvements in safety management methods. Since the introduction of smart construction technologies and stricter regulations contribute to accidents either during construction or after they have occurred, a safety management system capable of preventing accidents by identifying risk factors in advance is required.

[0007] Construction accidents occur not due to a single risk factor present at the site, but due to the complex interplay of various risk factors. Accordingly, methods for measuring the pre-risk level of construction sites by combining various risk factors have been studied. Existing research focuses on measuring pre-risk based on existing accident data and specific factors, such as calculating risk by combining work types, factors, and accident types, or presenting the magnitude of risk by combining weather conditions, worker characteristics, accident frequency, and accident intensity. While various factors exist as causes of accidents at construction sites, a risk assessment that reflects the factors most closely associated with fatal accidents is necessary to effectively reduce the number of fatalities. The problem to be solved

[0008] Currently, in the construction industry, in accordance with Article 31 of the Occupational Safety and Health Act, on-site workers must complete basic safety and health training at least once and submit a certificate of completion before being deployed to construction sites. This basic training currently consists of one hour on enhancing safety awareness, two hours on risk factors and safe work methods for each task, and one hour on health hazard risk factors and health management methods specific to construction occupations. In addition to this, site-specific safety training is conducted before work begins, but it is often carried out in a very simplified manner, rendering it ineffective. Furthermore, small and medium-sized construction companies frequently skip safety training due to tight project schedules. Consequently, the current state of safety education and the cultivation of safety awareness among construction workers in the industry is severely lacking. Generally, the construction industry is broadly classified into architecture, civil engineering, and plant construction. Since all three types have a higher probability of fatalities in the event of an accident compared to other industries, a proactive worker safety management system is required to prevent such incidents. While the most accurate approach would be to classify these three types, identify risk factors at every stage of each, and establish a worker safety management system optimized for that specific type, this presents difficulties in reality due to the limited availability of information. Therefore, in order to establish a practical and universal safety index evaluation model, the top five types of fatal accidents in the construction industry in 2021 were analyzed, and it was confirmed that falls (59.5%), collisions (8.9%), being struck by objects (7.2%), being crushed or overturned (6.2%), and collapses (6.0%) accounted for 87.8% of all fatal accidents (Ministry of Employment and Labor 2021). Accordingly, the purpose of this invention is to derive factors affecting worker safety at all construction stages of building, civil engineering, and plant construction sites that can reflect the top five types of fatal accidents in the construction industry in 2021, and to construct a construction site worker safety index evaluation model capable of quantitatively identifying the risk level of workers at each site by combining weights reflecting the correlations of the derived factors with the calculation of risk scales according to the items of the factors, and to propose a system and method for evaluating the safety index of construction site workers through this model.

[0009] The construction site worker safety index evaluation model proposed in this invention identifies risk factors that may lead to accidents during the construction phase and quantifies these factors using an Analytic Network Process (ANP) that reflects the interrelationships and feedback structures between factors. Through this model, it presents a quantified worker risk level that reflects changing working conditions during the construction phase and the individual characteristics of each site, thereby enabling the prior identification of workers with a high probability of risk from the initial to the final stages of construction. This allows for customized safety training for high-risk workers and tailored safety management by safety managers and site supervisors, ultimately contributing to the prevention of construction site safety accidents and the reduction of fatal accidents in the construction industry. means of solving the problem

[0010] According to one aspect of the present invention, a safety index evaluation system for a construction site worker is provided, comprising: a numerical model construction step for constructing a numerical model for evaluating the safety index of a construction site worker—wherein the safety index is determined by a plurality of predefined influencing factors, and the numerical model is defined as [Equation 1] below; an evaluation data acquisition step for the safety index evaluation system for acquiring evaluation data for each of the plurality of influencing factors of the worker to be evaluated; a risk scale determination step for the safety index evaluation system for determining a risk scale for each of the plurality of influencing factors of the worker to be evaluated based on the evaluation data for each of the plurality of influencing factors of the worker to be evaluated; and a safety index calculation step for the safety index evaluation system for applying the risk scale for each of the plurality of influencing factors of the worker to be evaluated to the numerical model to calculate the safety index of the worker to be evaluated.

[0011] [Formula 1]

[0012]

[0013]

[0014] (Here, n is the number of the aforementioned multiple influencing factors, V(a) is the risk index of worker a, w i is the weight of the i-th influencing factor, v i (a) is the risk scale of the i-th influencing factor of worker a, and S(a) is the safety index of worker a)

[0015] In one embodiment, the numerical model construction step further includes a weight determination step for determining weights for each of the plurality of risk factors through an Analytic Network Process (ANP) technique, wherein the plurality of influencing factors are divided into three influencing groups: a construction site group, an environmental group, and a worker group, and the plurality of influencing factors include: construction type, basic work type, site scale, construction period, and progress rate belonging to the construction site group; month of accident occurrence, time of accident occurrence, working layer, and temperature belonging to the environmental group; and age, work experience, occupation, and employment type belonging to the worker group, and the weight determination step comprises: a step of obtaining worker survey data for determining weights of the numerical model; and a step of performing pairwise comparisons between the three influencing groups, pairwise comparisons between influencing factors belonging to the construction site group, pairwise comparisons between influencing factors belonging to the environmental group, and pairwise comparisons between influencing factors belonging to the worker group based on the worker survey data. The method may include the step of generating a super matrix based on the results of the pairwise comparison and determining weights for each of the plurality of influence factors based on the generated super matrix.

[0016] In one embodiment, the plurality of influencing factors are classified into two categories: data-based impact classification and survey-based impact classification, and the plurality of influencing factors include site size, month of accident occurrence, time of accident occurrence, temperature, age, and work experience belonging to the data-based impact classification; and construction type, basic construction type, construction period, progress rate, work layer, occupation, and employment type belonging to the survey-based impact classification, and the numerical model construction step includes: a first risk scale determination step for determining the risk scale of each data-based influencing factor belonging to the data-based impact classification based on statistical data obtained from the database of the Korea Occupational Safety and Health Agency (KOSHA); and a second risk scale determination step for determining the risk scale of each survey-based influencing factor belonging to the survey-based impact classification based on expert survey data, and the first risk scale determination step includes, for each data-based influencing factor, a step of generating a frequency distribution table of the data-based influencing factor and the frequency of accident occurrence based on the statistical data. The method may include the steps of: sorting the classes of the above frequency distribution table in ascending order according to frequency and calculating the logarithm of the cumulative occurrence proportion; performing exponential regression analysis to calculate the coefficients of a utility function in the form of a natural logarithm; deriving a utility function defined by the calculated coefficients; and determining the risk scale of the data-based influencing factor based on the derived utility function.

[0017] According to another aspect of the present invention, a computer-readable recording medium is provided on which a computer program for performing the above-described method is recorded.

[0018] According to another aspect of the present invention, a computer program is provided that is installed in a data processing device and performs the method described above.

[0019] According to another aspect of the present invention, a safety index evaluation system for a construction site worker is provided, comprising at least one processor; and a memory in which a computer program is stored, wherein, when the computer program is executed by the at least one processor, the safety index evaluation system performs the method described above.

[0020] According to another aspect of the present invention, a safety index evaluation system for a construction site worker is provided, comprising: a storage module storing a numerical model for evaluating a safety index of a construction site worker—wherein the safety index is determined by a plurality of predefined influencing factors, and the numerical model is defined as in [Equation 1]; an evaluation data acquisition module for acquiring evaluation data for each of the plurality of influencing factors of the worker to be evaluated; a risk scale determination module for determining a risk scale for each of the plurality of influencing factors of the worker to be evaluated based on the evaluation data for each of the plurality of influencing factors of the worker to be evaluated; and a safety index calculation module for calculating a safety index of the worker to be evaluated by applying the risk scale for each of the plurality of influencing factors of the worker to be evaluated to the numerical model.

[0021] In one embodiment, the safety index evaluation system for a construction site worker further includes a weight determination module that determines weights for each of the plurality of risk factors through an Analytic Network Process (ANP) technique, wherein the plurality of influencing factors are divided into three influence groups: a construction site group, an environmental group, and a worker group, and the plurality of influencing factors include: a type of construction, basic work type, site scale, construction period, and progress rate belonging to the construction site group; and a month of accident occurrence, time of accident occurrence, working floor, and temperature belonging to the environmental group. and includes age, work experience, occupation, and employment type belonging to the above worker group, and the weight determination module acquires worker survey data for determining weights of the above numerical model, and based on the worker survey data performs pairwise comparisons between the three influence groups, pairwise comparisons between influence factors belonging to the above construction site group, pairwise comparisons between influence factors belonging to the above environmental group, and pairwise comparisons between influence factors belonging to the above worker group, generates a super matrix based on the results of the pairwise comparisons, and can determine weights for each of the plurality of influence factors based on the generated super matrix.

[0022] In one embodiment, the plurality of influencing factors are classified into two categories: data-based impact classification and survey-based impact classification, and the plurality of influencing factors include site size, month of accident occurrence, time of accident occurrence, temperature, age, and work experience belonging to the data-based impact classification; and construction type, basic construction type, construction period, progress rate, work layer, occupation, and employment type belonging to the survey-based impact classification, and the safety index evaluation system for construction site workers includes a first risk scale determination module that determines the risk scale of each data-based influencing factor belonging to the data-based impact classification based on statistical data obtained from the database of the Korea Occupational Safety and Health Agency (KOSHA); It includes a second risk scale determination module that determines the risk scale of each survey-based influencing factor belonging to the survey-based influencing classification based on expert survey data, and the first risk scale determination module can, for each of the data-based influencing factors, generate a frequency distribution table of the data-based influencing factor and the frequency of disaster occurrence based on the statistical data, sort the classes of the frequency distribution table in ascending order according to frequency, calculate the log value of the cumulative occurrence proportion, perform exponential regression analysis to calculate the coefficient of a utility function in the form of a natural logarithm, derive a utility function defined by the calculated coefficient, and determine the risk scale of the data-based influencing factor based on the derived utility function. Effects of the invention

[0023] This invention presents a model capable of quantitatively evaluating the risk level of workers by considering the conditions of factors affecting worker safety at construction sites. First, based on a review of existing research and accident data provided by KOSHA, a total of 18 risk factors were derived by classifying them into construction site, environmental, and worker groups. Regarding the derived 18 factors, a first survey was conducted among experienced safety management personnel working at clients, contractors, and engineering firms. Five factors that were not statistically significant were excluded, resulting in the final 13 risk factors (type of construction, basic work type, site scale, construction period, progress rate, month of accident, time of accident, work floor, temperature, age, length of service, occupation, and employment type). Subsequently, a second survey was conducted on the 13 risk factors to derive weights for each risk factor. A network was constructed by identifying correlations between influence groups and detailed factors through internal research team meetings and expert surveys. Using the ANP method, final weights were calculated for each group and risk factor, reflecting internal and external influences on the three groups and detailed factors. Next, to quantitatively evaluate the risk level of workers, a utility function-based risk scale was calculated to be combined with the final weights for each risk factor. The risk scale utilized actual construction site accident data provided by KOSHA (site size, month of accident, time of accident, temperature, age, length of service). For data where objective information was not available (type of construction, basic work type, construction period, progress rate, worker layer, occupation, employment type), a risk scale ranging from 1 to 10 was calculated for each detailed factor through expert surveys. Finally, a construction site worker safety index evaluation model in the form of a multi-criteria decision model was presented, capable of quantitatively evaluating worker safety by combining the final weights for each group and risk factor with the risk scales for each detailed factor.

[0024] Through the construction site worker safety index evaluation model presented in this invention, the risk level of workers is quantitatively calculated by reflecting variable factors at the construction site, thereby enabling the prior identification of workers with a high probability of risk occurrence based on site conditions at the pre-construction stage. This preemptive safety management system supports site-specific safety management by construction site safety managers, which is expected to reduce the number of fatal accidents in the construction industry and further contribute to improving the negative image of the industry caused by safety accidents. Brief explanation of the drawing

[0025] A brief description of each drawing is provided to help to better understand the drawings cited in the detailed description of the invention. FIG. 1 is a block diagram showing the structure of a safety index evaluation system for construction site workers according to one embodiment of the present invention. FIG. 2 is a diagram illustrating the process of a safety index evaluation system according to one embodiment of the present invention generating a prediction model and predicting the next time series data. FIG. 3 is a flowchart illustrating an example of the process of estimating parameters of a time series data prediction model in a safety index evaluation system according to an embodiment of the present invention. Figures 4a to 4g are drawings illustrating experimental results, respectively. Specific details for implementing the invention

[0026] The present invention is capable of various modifications and may have various embodiments; specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood that it includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the present invention. In describing the present invention, detailed descriptions of related prior art are omitted if it is determined that such detailed descriptions may obscure the essence of the present invention.

[0027] Terms such as "first," "second," etc., may be used to describe various components, but said components should not be limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0028] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. The singular expression includes the plural expression unless the context clearly indicates otherwise.

[0029] In this specification, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0030] Furthermore, in this specification, when one component 'transmits' data to another component, it means that the component may transmit the data directly to the other component or transmit the data to the other component through at least one other component. Conversely, when one component 'transmits' data directly to another component, it means that the data is transmitted from the component to the other component without passing through another component.

[0031] Below, we first examine the methodology necessary to understand the safety index evaluation system and method for construction site workers according to the technical concept of the present invention.

[0032] Methodology

[0033] 1. Multi-criteria Decision Making Method for Worker Safety Accident Evaluation

[0034] Multi-criteria Decision Making is a method that supports optimal decision-making using scientific means for complex problems involving multiple criteria, such as when management issues arise in corporations or public institutions, or when decisions need to be derived from an overall perspective. Most industries utilize Multi-criteria Decision Making to derive decisions, including selecting methods for analyzing potential process risks at manufacturing sites for hazardous material handling machine parts, analyzing the causes of machine accidents and devising countermeasures, presenting methods for selecting weapon systems by deriving priorities, and deriving criteria to evaluate smart grid-related technologies and services to ultimately determine priorities for alternatives. Although it is difficult to clearly distinguish the types of Multi-criteria Decision Making, they can be classified into Multiple Objective Decision Making (MODM) and Multiple Attribute Decision Making (MADM) depending on whether the alternatives are infinite or finite. Since complex problems occurring at construction sites generally exhibit characteristics involving a finite number of alternatives, this study intends to adopt the Multiple Attribute Decision Making method most suitable for this invention by reviewing cases where MADM has been applied to the construction industry.

[0035] 1.1 Multi-Attribute Utility Theory (MAUT)

[0036] The multi-attribute utility function method, which originated from the hypothesis of Von Neumann and Morgenstern that different preferences exist among the alternatives available in the problem of uncertainty, is generally used to make a rational choice among alternatives that include multiple attributes. This method first derives individual attribute utility functions to evaluate and combine the utility of each attribute, and then calculates a multi-attribute utility function through their combination. The general utility function equation for cases where mutual utility independence holds is given by Equation (1) below.

[0037]

[0038] Elpidio Oscar Benitez Nara (2019) developed the Occupational Health and Safety (OHS) method to derive individual competitiveness rates for companies and combined this with a Neural Network to derive Key Performance Indicators (KPIs) that have the greatest impact on corporate competitiveness. Hatush (1998) used MAUT to establish an evaluation system focusing on the adoption of assembly methods and assembly optimization to evaluate concrete building construction methods during the design phase, and Ying Chen (2010) evaluated the feasibility of pre-assembly in the early stages of a project based on MAUT to derive the optimal strategy for applying pre-assembly in concrete buildings. Casanovas-Rubioetal (2018) presented a decision model utilizing MAUT to evaluate labor risk assessment methods, related technologies, and sustainability in post-disaster housing construction. While MAUT has the advantage of quantifying qualitative attributes to determine priorities, it has the disadvantage that as the number of qualitative attributes increases, deriving the utility function becomes complex and verifying the consistency of responses to those qualitative attributes is difficult.

[0039] 1.2 Scoring Model (Scoring Method)

[0040] To determine the priority of alternatives, the method assigns relative weights to each factor, following the procedure of weighting factors, deriving factor-specific scores for each alternative, and selecting priorities through the comparison of comprehensive scores. Detailed methods for calculating weights are broadly classified into three types: the simple weighting method (dividing by the number of factors), the rank sum weighting method (assigning weights based on the reverse order of rankings), and the rank inverse weighting method. Among these three methods, the rank inverse weighting method is generally used, and the weight calculation method is as shown in Equation 2 below. Cheung (2004) built a web-based Construction Safety and Health Monitoring (CSHM) system that evaluates health performance factors using the Scoring Method. Oleg Kaplinski (2008) proposed an evaluation method capable of verifying a company's financial status by classifying the priorities of five types of factors: turnover capital, retained profit, profit before tax and interest repayment, market value of share capital, and sales income. While the scoring model has the advantage of being able to evaluate alternatives by calculating weights very intuitively and simply compared to other methods, it has the disadvantages of being subjective and difficult to verify the consistency of rankings.

[0041]

[0042] 1.3 Outranking Method

[0043] This is an analytical method that uses a dichotomous relationship by creating pairs to compare two alternatives and determine which one is more suitable. A representative technique is the Elimination and Choice Translating Relation (ELECTRE) method, which follows five main steps. It consists of: 1) normalization of alternatives of different scales; 2) multiplication of the normalized values ​​by weights; 3) comparison of superiority and inferiority of each alternative; 4) calculation of the concordance index and discordance index; and 5) elimination of alternatives using a threshold value. The method for calculating the concordance and discordance indices is shown in Equations 3 and 4 below (Korea Development Institute 2000). Chou (2007) quantified the ecological, safety, and cost indices of construction sites using fuzzy theory, linked the ecological indices, and then proposed an optimal ecological technology method using the PROMETHEE method. Liu (2019) constructed an evaluation model that supports decision-making regarding sustainability for designers and developers by combining AHP and Fuzzy ELECTRE techniques. Park (2008) presented a model that evaluates flood risk by classifying it into hydrological, topographic, and human characteristics, enabling the establishment of priority flood mitigation measures in areas expected to be flooded. The Outranking Method has the advantage of being highly diverse in its techniques and allowing for detailed qualitative and quantitative analysis using indices. However, it has disadvantages such as ambiguous threshold values, difficulty in selecting alternatives as the number of alternatives increases, and difficulty in applying it to similar projects other than the project in question.

[0044]

[0045]

[0046] 1.4 Analytic Hierarchy Process (AHP)

[0047] This technique supports decision-making by systematically evaluating mutually exclusive alternatives to derive priorities. It hierarchizes the various evaluation factors constituting a problem into major and minor factors, and derives weights for each factor by performing pairwise comparisons to indicate relative importance and preference among the hierarchical factors. Generally, the process follows the steps of: 1) identifying the decision target and conceptualizing the evaluation; 2) establishing evaluation criteria and a hierarchical structure; 3) calculating weights for each criterion; 4) calculating scores for each criterion item; and 5) calculating a comprehensive score. In step 4, calculating scores for each criterion item, scores are determined by assigning points to item weights using a utility function. Since not all evaluation criteria generally hold equal importance, the step of calculating item weights based on specific evaluation levels obtained through surveys or similar methods is critical. Aviad Shapira (2009) and Gabriel Raviv (2016) evaluated the relative importance of factors that can affect tower crane safety at construction sites based on weights calculated using AHP. Chan (2004) used AHP to analyze the differences in importance of items in the Hong Kong construction industry's occupational health and safety management system according to company type, and showed that the importance of factors differed for joint ventures (JV), well-established enterprises (WE), and small and medium-sized enterprises (SME). Zhang (2020) constructed a hierarchical safety assessment system capable of comprehensively evaluating risk factors affecting road construction by combining AHP and principal component analysis (PCA). The AHP technique allows for the measurement of preferences for objective factors as well as subjective and qualitative factors, and its evaluation system is universally applicable. Furthermore, it is frequently used in the construction industry due to the advantage of being able to derive reliable results even with a relatively small sample size.However, it is necessary to construct a scale that ensures respondent consistency while allowing preferences to be easily expressed, and caution is required as errors may occur during the weight calculation process depending on the evaluation item design method. Additionally, there is a disadvantage in that calculations based solely on unidirectional flow are possible because evaluation items and factors are treated independently without considering correlations.

[0048]

[0049] 2. Analytic Network Process (ANP)

[0050] ANP is an extended concept of AHP and is a multi-criteria decision-making technique that performs sophisticated decision-making by reflecting the interrelationships and feedback structures among evaluation factors. As shown in <Figure 1>, unlike the AHP hierarchical structure (or downward hierarchical tree structure) which vertically stratifies a problem and considers detailed factors as mutually independent, ANP incorporates cycles and loops among detailed factors that have interrelationships. In the ANP model, the basic unit of the network is a cluster, and each cluster contains multiple decision factors (nodes). Mutual influence relationships can be identified based on the assumption that factors belonging to a specific cluster can influence factors belonging to other clusters. It differs from AHP in that it recognizes internal and external dependencies among various criteria and calculates weights by representing them as a macromatrix.

[0051] The construction site worker safety index model intended to be derived in this invention measures risk based on factors that may affect workers during the construction phase, making it difficult to quantitatively evaluate the importance of such factors. Furthermore, given the existence of correlations and interdependencies among the factors, it was determined that using ANP, which can reflect these factors, would allow for the derivation of more accurate relative importance.

[0052] The application process of ANP to calculate the weights of risk groups and factors related to worker safety during the construction phase at a construction site is divided into four steps: 1) building a network model, 2) pairwise comparison based on the influence criteria of each factor and generation of preference vectors, 3) generation and transformation of a supermatrix, and 4) calculation of a comprehensive score based on the relative importance of each factor and final preference. Vector values ​​are derived as shown in Equation (5) below through pairwise comparisons of groups and factors corresponding to the entire network. When performing pairwise comparisons, it is necessary to verify the consistency of the survey evaluations, and consistency verification is considered satisfied if the Consistency Ratio (CR), which is the ratio of the Consistency Index (CI) to the Random Index (RI), is 0.1 or less.

[0053]

[0054] A super matrix is ​​constructed for the factors that passed the consistency check. Basically, the super matrix consists of m clusters, C h Assuming there is a network composed of (h=1,2,3···,m), each cluster is n h It has decision factors, and this is e h1 , e h2 , e h3 , … , e hm It is expressed as such. In this case, the weights constituting the super matrix are obtained through pairwise comparisons between factors, and the weight w ij represents the degree of influence that attribute i has on attribute j, and is as shown in equation (6) below.

[0055]

[0056] To calculate weights, three types of super matrix processes are performed: unweighted super matrix, weighted super matrix, and limited super matrix. First, an unweighted super matrix is ​​derived by calculating weights based on the results of pairwise comparisons of clusters. Subsequently, a weighted super matrix is ​​derived by multiplying the weights obtained from node pairwise comparisons by the cluster weights, reflecting the correlation between the upper and lower factors respectively, with the sum of each column equal to 1. Finally, the weighted super matrix is ​​raised to a power until it converges to a constant value, and a limited super matrix is ​​constructed based on this constant value.

[0057] 3. Establishment of Utility Measures by Risk Level Assessment Factor for Construction Site Workers

[0058] 3.1 Extraction of Utility Measures by Detailed Factors Based on Data and Expert Experience

[0059] A utility function is generally a function that represents the relationship between actual values ​​and utility, signifying a quantification of decision-making attitudes toward uncertainty. A single-attribute utility function can be derived by quantifying the utility and preference concepts of individual attributes related to evaluation items, and a multi-attribute utility function can be calculated through their combination. The form of a general utility function can be determined as follows, depending on the risk attitude (risk-neutral, risk-averse, risk-preferred) toward the attributes.

[0060]

[0061] Here, risk neutrality refers to the case where the certainty equivalent—that is, a certain value that the decision-maker feels is indifferent to any uncertain event—is regarded as being on par with the expected value; this describes a situation where decisions are made solely based on the expected value, without reflecting the decision-maker's attitude toward risk. Risk adverseity refers to the case where the certainty equivalent is viewed as smaller than the expected value, signifying a cautious approach to risk. Risk proneness refers to the attitude of viewing the certainty equivalent as larger than the expected value, seeking to maximize potential gains even while assuming risk. The utility function for a risk-averse tendency is represented as a concave function, the utility function for a risk-seeking tendency as a convex function, and the utility function for a risk-neutral tendency as a linear function.

[0062] In the present invention, a single-attribute utility function for individual evaluation factors is derived by referring to the application procedure of a general multi-attribute decision model, and a risk scale for each attribute based on the utility function is calculated, and the level of safety accidents or safety index of construction site workers is calculated through a combination of these and weights for each factor.

[0063] If a quantitative database (DB) related to evaluation items exists for individual attributes, data with different units is standardized, and a utility function—that is, a risk scale for each attribute—is derived from the normalized data. On the other hand, for attributes where quantitative data is absent, a utility function for each goal is derived from the responses of experts. In this way, when quantifying the scale of a worker's risk level based on the experience of an expert group, the utility function is derived by assuming a risk-neutral propensity among the three risk attitudes mentioned above, taking into account the average decision-making propensity of the experts.

[0064] 3.2 Construction Site Worker Safety Index Evaluation Model (Worker Safety Index)

[0065] The multi-criteria decision-making model to be developed in this study combines the advantages of (1) the ANP method, which reflects the correlation and interdependence between each component factor when calculating weights for each evaluation item, and (2) the advantages of the Multi-attribute Utility Theory (hereinafter MAUT), which directly measures the utility of the decision-maker. Here, MAUT refers to a theory that supports problem solving by subdividing the evaluation target into various attributes and quantifying the decision-maker's value, utility, and preference levels for individual attributes. The core of MAUT is to measure the preference for a specific goal or alternative, and a standard score is assigned to the degree of utility provided by each evaluation item or alternative based on the decision-maker's subjective preference level for each evaluation criterion or objective data.

[0066] Accordingly, in this study, influence groups and detailed factors affecting the safety of construction site workers are first identified through a review of existing literature and a survey. The relationships in which influence groups and detailed factors exert internal and external influences are identified, and finally, the final weights of the influence groups and detailed factors based on the ANP method are calculated by reflecting the correlations between factors. Subsequently, a risk scale of 10 points is derived for each item of the detailed factor through objective data and a survey of practitioners. Once the weights of each influence group and detailed factor based on ANP and the risk scales for the items of the detailed factor based on utility functions are derived, the weights of each influence group and detailed factor are multiplied by the risk scale corresponding to each worker, and the values ​​are summed to produce a final score that quantifies the degree of risk of the worker, which is given by the following equation (10).

[0067]

[0068] Proposed Worker Safety Index Evaluation Model

[0069] 1. Framework for Developing a Worker Safety Index Evaluation Model

[0070] The procedure for developing the worker safety index evaluation model is as shown in Figure 1. First, factors affecting worker safety are identified through a review of existing literature and data provided by the Korea Occupational Safety and Health Agency (KOSHA). Subsequently, a first survey was conducted with experts experienced in safety management regarding the identified influencing factors, and the final influencing factors were derived by verifying the statistical significance of the survey results. Since the weights of the influencing factors to be applied to the worker safety index evaluation model must reflect the correlations between groups and detailed factors, a network between the influencing groups of the construction site worker safety index is first constructed to identify these correlations. Based on the constructed network, a second survey is conducted to identify the importance of each factor; when the results of the second survey evaluation are derived, only those with a response consistency ratio of less than 0.2 are accepted, and a grand matrix is ​​calculated using the ANP method. Through the calculated grand matrix, the final weights of the 13 influencing factors, reflecting the correlations between groups and influencing factors, are derived. Subsequently, risk scales for detailed factor items are derived based on data from 61,818 accident cases provided by KOSHA and results obtained through surveys. Finally, a construction site worker safety index evaluation model in the form of a multi-criteria decision model is presented by combining the previously derived weights and detailed factor risk scales.

[0071] 2. Survey Overview

[0072] In the research to derive the present invention, a survey was conducted by classifying respondents into clients, construction companies, and engineering firms to develop a construction site worker safety index evaluation model that reflects opinions from various perspectives regarding construction site safety. The survey subjects were selected from experienced safety management personnel working at clients, construction companies, and engineering firms, and two surveys were conducted to evaluate the suitability of factors affecting worker safety, calculate weights for verified factors, and derive risk scales for detailed factor items. The first survey was conducted on 11 clients (16%), 44 construction companies (65%), and 13 engineering firms (19%). The experience of the respondents was classified as follows: less than 5 years (19 people, 28%), 5 to 10 years (8 people, 12%), 10 to 15 years (9 people, 13%), 15 to 20 years (8 people, 12%), and 20 years or more (24 people, 35%). The second survey was conducted with 12 clients, 39 contractors, and 18 engineering firms; among them, 5 clients (17%), 17 contractors (59%), and 7 engineering firms (24%) satisfied the consistency index (0.2 or less). Regarding the experience of the respondents who satisfied the consistency index, 7 (24%) had less than 5 years, 6 (21%) had 5–10 years, 2 (7%) had 10–15 years, 4 (14%) had 15–20 years, and 10 (34%) had 20 years or more. Since the proportion of respondents with 15 years or more of experience was 47% in the first survey and 48% in the second, the survey results were deemed representative. For the first survey, risk factors were identified through a review of existing literature, and the significance of these factors was determined using a 10-point Likert scale evaluation. Although 4- to 7-point Likert scales are generally widely used for reasons of response advantage (Shing-On Leung, A comparison of Psy...), in this study, a survey was conducted using a 10-point scale for the speed and ease of worker judgment and for intuitive perception when converting the final model to a 100-point scale.

[0073] 3. ANP-based weight calculation

[0074] 3.1 Derivation of Key Factors for Measuring Worker Risk Levels

[0075] In this invention, factors affecting accidents during the construction phase of a construction project are first classified into three main groups—construction site group, environmental group, and worker group—through the analysis of data from the Korea Occupational Safety and Health Agency (KOSHA) and existing domestic and international studies that have calculated construction site risks. The construction site group includes six detailed factors: type of construction, basic work type, construction cost, site scale (number of employed workers), construction period, and progress rate; the environmental group includes the month of accident occurrence, day of the week of accident occurrence, time of accident occurrence, working floor, temperature, and precipitation; and the worker group includes six detailed factors: gender, age, work experience (length of service), occupation, safety training hours, and employment type. Thus, a total of 18 detailed factors are derived. A first survey was conducted to evaluate the importance of the 18 derived detailed factors using a 10-point Likert scale, targeting 68 experts in construction projects and construction safety, including architectural construction, construction management, and site safety management. As shown in , the p-value of each detailed factor derived based on the t-test is verified to be statistically significant. Although most influencing factors were significant at the 95% confidence level, five detailed factors—construction cost within the construction site group, day of the week and precipitation within the environmental group, and gender and safety training hours within the worker group—were excluded as they were not statistically significant. Finally, a total of 13 factors were derived that directly affect worker safety during the construction phase: five from the construction site group (type of construction, basic work type, site scale (number of employed workers), construction period, progress rate), four from the environmental group (month of accident, time of accident, working floor, temperature), and four from the worker group (age, work experience (length of service), occupation, employment type).To construct the super matrix used for deriving weights, the construction site group among the influence groups is denoted as C, with detailed factors such as type of construction C1, basic type of work C2, site size (number of employed workers) C3, construction period C4, and progress rate C5; the environmental group is denoted as E, with detailed factors such as date of accident (month) E1, date of accident (hour) E2, work floor E3, and temperature E4; and the worker group is denoted as W, with detailed factors such as age W1, work experience (length of service) W2, job type W3, and employment type W4, which are summarized in .

[0076]

[0077]

[0078] 3.2 Worker Safety Index Evaluation Model Network

[0079] Factors affecting worker safety at construction sites are highly likely to be interrelated. Therefore, it is crucial to identify the degree of influence among these influencing factors to derive accurate weights. To derive weights that reflect the interrelationships between evaluation factors using the ANP technique, it is first necessary to establish a network capable of determining the influence between groups and sub-factors, which allows for identifying which factors influence others. Sub-factors belonging to the three influence groups (construction site group, environmental group, and worker group) identified in this invention were determined to exert internal influence. Regarding external influence between influence groups, expert consultation revealed that there is mutual influence between the construction site group and the environmental group, as well as between the worker group and the environmental group. An example of the degree of influence between influence groups involves determining which factor—such as the month of accident occurrence in the environmental group or the occupation and employment type in the worker group—has the greatest impact. Similarly, the degree of influence within an influence group involves determining which factor—such as the site scale (number of employed workers) in the construction site group or the construction period or the completion rate—has the greatest impact. Accordingly, the network between influence groups presented in the present invention was structured as shown in Fig. 2.

[0080] 3.3. Calculation of Key Factor Weights

[0081] The 13 detailed factors derived earlier are qualitative factors that must be converted quantitatively to calculate weights. Accordingly, regarding the final 13 detailed factors derived from the results of the first survey, a second survey was conducted to perform pairwise comparisons of influence by influence group, by detailed factor item within each group, and between influence groups and detailed factors. The second survey was conducted with a total of 69 participants. Generally, results with a consistency ratio of less than 0.1 are accepted, but since inconsistency up to less than 0.2 is considered within an acceptable range, it was decided to accept survey results with a consistency ratio of less than 0.2 for the study to derive this invention. The survey participants with a consistency ratio of less than 0.2 consisted of 5 clients, 17 construction companies, and 7 engineering firms; factor-specific weights were derived based on the responses of 29 out of the 69 participants. Pairwise comparisons were performed between the influence group and the sub-factors within the influence group using the geometric mean of the pairwise comparison results of the 29 questionnaires, and the consistency indices were 0.0048, 0.001, 0.0004, and 0.0002, respectively, indicating consistency in the responses.

[0082] The results of the pairwise comparison analysis among the influence groups are summarized in , showing large weights in the order of the worker group (0.474), construction site group (0.314), and environmental group (0.211). The results of the pairwise comparison analysis among the detailed items of the construction site group within the influence groups are summarized in , showing relatively large values ​​for construction type (0.232) and site scale (0.205). The results of the pairwise comparison analysis among the detailed items of the environmental group are summarized in , showing relatively large values ​​for working layer (0.316) and temperature (0.257), and as shown in , among the detailed items of the worker group, job type (0.335) and work experience (0.284) show large values. The weight results derived through pairwise comparison are used to calculate the final weights through the calculation of internal and external influences.

[0083]

[0084]

[0085]

[0086]

[0087] To obtain weights reflecting the influence between evaluation factors using the ANP technique, an unweighted super matrix is ​​first constructed as shown in using the previously calculated weights of the influence group and factors within the influence group, as well as the influence weights between the inside and outside of the influence group. Subsequently, the derived unweighted super matrix and cluster weights are combined to construct a weighted super matrix such that the sum of each column is 1, which is summarized in . To calculate the final weights, the weighted super matrix is ​​repeatedly raised to powers until it converges to a constant value (n=9), which is the final weight to be obtained in one embodiment of the present invention, and is summarized in .

[0088]

[0089]

[0090]

[0091] The final weights of construction site risk factors are calculated by reflecting the influence between evaluation factors using a super matrix. Engineering firms were classified into three groups—Client & Engineering Firm, Construction Firm, and All Subjects—assuming that their primary role is construction site supervision and thus they share a similar perspective with the client. Both Client & Engineering Firm and Construction Firm showed the highest weight for the Environmental Group; for Client & ENG Firm, the Construction Site Group had a slightly higher weight than the Worker Group, while for Construction Firm, the weights were equal.

[0092] When examining the weights of factors affecting the safety of construction site workers across the entire pool of clients, contractors, and engineering firms, the impact groups are ranked in the order of the Environmental Group (0.427), Worker Group (0.284), and Construction Site Group (0.283). Safety accidents occurring at construction sites are most significantly influenced by the working environment; the Worker Group, which includes age, work experience, occupation, and employment type, has the second greatest impact, while the Construction Site Group, which includes construction type, basic work category, and construction period, has the least impact on worker safety. Regarding specific factors, within the Environmental Group, the time of accident occurrence (0.120), which represents the worker's working hours, has the greatest influence, followed by the month of accident occurrence (0.116), temperature (0.096), and working floor (0.095). In the worker group, the occupation based on the worker's duties, such as carpentry, rebar work, and temporary construction, has the greatest impact at 0.079, followed by work experience (0.073), age (0.072), and employment type (0.060). Finally, in the construction site group, the type of construction, such as housing, factories, and roads, has the greatest impact at 0.071, while factors such as site size (0.054), construction period (0.054), basic construction type (0.052), and completion rate (0.052) have almost similar impacts, as summarized in below.

[0093]

[0094] 4. Risk assessment based on KOSHA data and practitioner experts

[0095] In the study for this invention, risk scales were derived for six of the 13 influencing factors on safety accidents among construction site workers (month of accident, time of accident, temperature, age, work experience (length of service), and site size (number of workers)) by utilizing cases provided by KOSHA. Although the statistical data provided by KOSHA contains information on 64,744 construction safety accidents that occurred between 2013 and 2018, information was not disclosed at the same level for all items; therefore, a data preprocessing process was performed prior to data analysis. On the other hand, for influencing factors for which objective statistical data has not been systematically accumulated (type of construction, basic construction type, construction period, progress rate, worker layer, occupation, and employment type), risk levels for each factor were derived through a survey of a group of practical experts.

[0096] 4.1. Derivation of Risk Measures Using Objective Databases (DB)

[0097] In the study for the present invention, risk levels are calculated using case-related statistical data provided by KOSHA for 6 of the 13 major factors affecting safety accidents among construction site workers (date and month of accident, time of accident, temperature, age, work experience (length of service), and site size (number of workers)).

[0098] Among the detailed factors belonging to the 'Construction Site Group' within the influence group, the factor for which a risk scale was derived using the KOSHA database is 'Site Size (Number of Employed Workers).' As a result of analyzing approximately 60,000 cases released by KOSHA regarding the risk scale of this factor, it was determined that 'sites with fewer than 5 employees' would have the highest risk scale related to the 'Site Size (Number of Employed Workers)' factor. Among the detailed factors belonging to the 'Environmental Group,' the factors for which a risk scale was derived using the KOSHA database are 'Month of Accident,' 'Time of Accident,' and 'Temperature.' Using cases released by KOSHA, the cumulative probability of occurrence frequency was calculated as shown in , and the groups were reorganized in ascending order. The cumulative weights were then converted to logarithmic values ​​to apply a linear regression model. Among the detailed factors belonging to the 'Worker Factor,' 'Age' and 'Length of Service' were also logarithmetized by grade in the same manner, and linear regression analysis was performed.

[0099]

[0100] The constants α and β and the risk avoidance coefficient r were derived from the three representative forms of utility function equations. Since information on evaluation attributes can be obtained and analyzed from actual accident cases, risk propensity was assumed to be risk-averse or risk-preferred; this implies that the utility function for the corresponding factor is derived in the form of an exponential equation based on risk avoidance or preference, rather than a linear equation based on risk neutrality. Accordingly, the utility function equation for the aforementioned six attributes is u(x) = βe ±rx To derive the form, exponential regression is performed as follows.

[0101] The first step is data analysis and natural logarithm transformation. For exponential regression analysis, it is necessary to define the causal variable (x) and the response variable (y). For the six major attributes, the probabilities corresponding to the data classes and ranges established to derive the frequency distribution table are rearranged in ascending order to set each group as the causal variable, and the probabilities corresponding to the classes for each factor are set as the response variable. Additionally, the natural logarithm of the response variable (ln(y)) is calculated for the exponential regression analysis. The results of the regression analysis are shown in below, and following general regression analysis procedures, the results can be interpreted in four main ways. First, the statistical significance of the regression analysis itself can be confirmed based on the F-value; if the 'significant F' value is lower than 0.05, the analysis result is interpreted as statistically significant. Second, the goodness of fit of the estimated model can be examined based on the 'coefficient of determination' and the 'adjusted coefficient of determination'. Here, the coefficient of determination ranges from 0 to 1 depending on the correlation between the dependent variable and the independent variable; the higher the correlation, the closer the value is to 1, which is interpreted as indicating higher utility of the regression model. Third, statistical significance can be confirmed based on whether the P-value for each individual independent variable is less than 0.05. Fourth, the regression equation, which is the utility function intended for use in this invention, can be derived using the 'y-intercept (β0)' and 'x1 (β1)' values ​​in . Considering this process, the F-value, the coefficient of determination, the adjusted coefficient of determination, and the P-value can all be interpreted as indicators verifying the utility and validity of the finally derived exponential regression equation.

[0102]

[0103] Looking at the results of the regression analysis above, it can be confirmed that all analysis models for the six attribute information are statistically significant in terms of significant F-values, coefficients of determination, adjusted coefficients of determination, and P-values. Accordingly, using the coefficients (β0, β1) in , a utility function equation as shown in is derived.

[0104]

[0105] Utility scales or risk scales are established as shown in by utilizing the utility functions of the six detailed factors. The scale ranges for the time of accident occurrence, temperature, age, etc., are defined as 1 to 10 points; the month of accident occurrence is defined as 5 to 10 points; and the site size and length of service are defined as 2 to 10 points and 3 to 10 points, respectively. The minimum value of the utility scale is the value calculated from the utility function, and if the difference between scales is small, it is defined as the same scale.

[0106]

[0107] 4.2. Derivation of a Survey-Based Single-Attribute Utility Function

[0108] For influencing factors (type of construction, basic construction type, construction period, progress rate, work layer, occupation, employment type) for which information was not provided in the KOSHA DB, the risk levels of experts were investigated according to the classification of factor items by detailed factor. For items evaluated by experts as having a high risk of safety accidents according to the classification of individual factor items, the highest score of 10 points was assigned, and a risk scale was calculated based on the ratio of the risk of other items to that item.

[0109] First, among the detailed factors corresponding to the impact group 'construction site,' the factors for which risk scales were derived based on expert surveys were 'type of construction,' 'basic construction type,' 'construction period,' and 'completion rate,' and the risk scales for these factors were calculated based on the risk levels of the experts. As a result of the analysis, regarding the 'type of construction' factor, the impact of safety hazards in 'railway, airport, and dam' construction was found to be higher compared to other construction projects, so the highest score of '10 points' was assigned to this category. For other category classifications ('single-family and multi-family housing,' 'Type 1 and Type 2 neighborhood living facilities,' etc.), risk scales were calculated by comparing them with the impact of safety hazards in 'railway, airport, and dam,' to which the highest score was assigned. Regarding the 'Basic Work Type' factor, as the impact of safety hazards on the 'Demolition and Renovation Work' type was found to be the highest compared to other types, the highest score of '10 points' was assigned to this category. For other category classifications ('General, Common, and Temporary Works', etc.), the risk scale was calculated by comparing it with the safety hazard impact of 'Demolition and Renovation Work'. Regarding 'Construction Period', the construction periods at sites where safety hazards occurred were classified from a minimum of 'less than 12 months' to a maximum of '36 months or more'. Since the analysis showed the highest frequency of safety hazard impacts for sites with a duration of '12 months or less', the highest score of '10 points' was assigned to this category. For other category classifications ('12–24 months', '24–36 months', '36 months or more'), the risk scale was calculated by comparing it with the safety hazard impact of sites with a duration of 'less than 12 months', to which the highest score was assigned. Finally, 'completion rate' is a factor representing the completion rate at the time of the accident at the site where the safety accident occurred. As it was analyzed that the impact of safety accidents was highest for a completion rate of '60%~80%', the risk scale for this item was calculated as 10 points. The risk scales for other item categories ('0%~20%', '20%~40%', ..., '80%~100%') were calculated differentially by comparing them with the impact score for the '60%~80%' completion rate.

[0110] Among the detailed factors corresponding to the impact group 'Environmental Factors,' the detailed factor for which a risk scale was derived through a survey is 'Work Floor,' and the risk scale for this factor was also calculated based on the risk levels of experts. The 'Work Floor' factor represents the work floor at the time of the accident at the site where the safety accident occurred; as the impact of safety accidents on the '30th floor or higher' work floor was analyzed to be the highest compared to other work floors, the risk scale for this item was set to 10 points. The risk scales for other item classifications ('1st floor or lower,' '2nd to 9th floors'... '20th to 29th floors') were calculated differentially by comparing them with the impact score for the '30th floor or higher' work floor.

[0111] Finally, among the detailed factors corresponding to the impact group 'Worker Factor,' the factors for which risk scales were derived through a survey are 'Occupation' and 'Employment Type.' The 'Occupation' factor indicates the occupation of workers affected by safety accidents; as the analysis showed that the impact of safety accidents on workers in the 'Scaffolder' occupation was higher compared to other occupations, the risk scale for this item was set to 10 points. The risk scales for other category classifications ('General Laborer / Contract Worker,' 'Carpenter,' ... 'Roofer') were calculated differentially by comparing them with the impact score for the 'Scaffolder' occupation. The 'Employment Type' factor was classified into only two categories—regular and irregular—so 'irregular workers,' who have a higher impact, were assigned a score of 10 points, while 'regular workers,' who have a relatively lower impact, were assigned a score of 5 points.

[0112] As a result of the analysis, risk scales by score interval and detailed factor classification were derived as shown in below.

[0113]

[0114] 5.5 Construction Site Worker Safety Index Evaluation Model

[0115] In the previous section, a safety index evaluation model for construction site workers was developed by utilizing the ANP technique to reflect the influence between the impact group and the detailed factors within the impact group, and by using objective actual accident data of 61,813 cases provided by KOSHA for the month of accident, time of accident, temperature, age, work experience (length of service), and site size (number of workers) for the six detailed factors, and by deriving risk scales based on risk levels obtained through surveys for the type of construction, basic construction type, construction period, progress rate, work layer, occupation, and employment type not provided by KOSHA. The proposed evaluation model is as shown in <Equation 11> below. It calculates a risk index for construction site workers out of 100 points by multiplying the weights of variable construction site groups (type of construction, basic work type, site scale, construction period, progress rate), environmental groups (date and month of accident, date and time of accident, working floor, temperature), and worker groups (age, work experience, occupation, employment type) by risk scales for detailed factor items, summing the results, and multiplying by 10. Finally, as shown in <Equation 12>, the previously derived construction site worker risk index is subtracted from 100 to calculate the final worker safety index. This enables quantitative evaluation by calculating the degree of risk of construction site workers as an index, and allows for customized special management by identifying high-risk workers and the prior identification of risk levels at each site through comparison with the overall worker index.

[0116]

[0117] Proposed System and Method (System and Method for Evaluating Safety Index of Construction Site Workers)

[0118] Below, the previously mentioned " Proposed Worker Safety Index Evaluation Model This section explains the safety index evaluation system and method for construction site workers using the worker safety index evaluation model described in the previous section.

[0119] FIG. 3 is a block diagram showing the structure of a safety index evaluation system for construction site workers (hereinafter referred to as the "safety index evaluation system") according to one embodiment of the present invention.

[0120] The above safety index evaluation system (100) may be a computing system, a server, or a data processing device. In one embodiment, the safety index evaluation system (100) may be a distributed processing system. Alternatively, the safety index evaluation system (100) may be a processing device including a desktop computer or a laptop computer, or a handheld device such as a mobile phone, a smartphone, a tablet PC, or a PDA (Personal Digital Assistant).

[0121] Meanwhile, the above safety index evaluation system may be equipped with hardware resources and / or software necessary to implement the technical concept of the present invention, and does not necessarily mean a single physical component or a single device. That is, the above safety index evaluation system may mean a logical combination of hardware and / or software provided to implement the technical concept of the present invention, and if necessary, it may be implemented as a set of logical configurations to implement the technical concept of the present invention by being installed in devices spaced apart from each other and performing respective functions. In addition, the above safety index evaluation system (100) may mean a set of configurations implemented separately for each function or role to implement the technical concept of the present invention. The above safety index evaluation system may be implemented in the form of a plurality of modules.

[0122] In this specification, the term "module" may refer to a functional and structural combination of hardware for carrying out the technical concept of the present invention and software for driving said hardware. For example, said module may refer to a logical unit of a specific code and a hardware resource for executing said code, and it can be easily inferred by an average expert in the technical field of the present invention that it does not necessarily refer to physically connected code or to a single type of hardware.

[0123] As illustrated in FIG. 3, the safety index evaluation system (100) according to one embodiment may include a storage module (110), an evaluation data acquisition module (120), a risk scale judgment module (130), a safety index calculation module (140), a construction module (150), and an output module (160), and the construction module may include a weight determination module (151), a first risk scale determination module (152), and a second risk scale determination module (153). According to an embodiment of the present invention, some of the components described above may not necessarily be components essential for the implementation of the present invention, and according to an embodiment, the safety index evaluation system (100) may include more components than those described above. For example, the safety index evaluation system (100) may further include a control module (not shown) for controlling the functions and / or resources of the components of the safety index evaluation system (100) (e.g., the storage module (110), the evaluation data acquisition module (120), the risk scale judgment module (130), the safety index calculation module (140), the construction module (150), the output module (160)) or a communication module (not shown) for communicating with an external device via a network.

[0124] Referring to FIG. 3, the storage module (110) can store a numerical model for evaluating the safety index of a construction site worker in advance. The storage module (110) can store the numerical model in a memory device equipped with the safety index evaluation system (100). The numerical model can be built in advance by the construction module (150) to be described later and stored by the storage module (110).

[0125] The above safety index may be determined by a plurality of predefined influencing factors. The safety index may be a measure or numerical value indicating how safe construction site workers subject to evaluation are from accidents, and each of the aforementioned plurality of influencing factors may be various factors analyzed to have a significant impact on the safety index. The influencing factors are the previously explained " Proposed Worker Safety Index Evaluation Model It is identical to the concept expressed in the part using terms such as detailed factors and key factors.

[0126] Meanwhile, the above numerical model can be defined as shown in [Equation 1] below.

[0127] [Formula 1]

[0128]

[0129]

[0130] Here, n is the number of the plurality of influencing factors, V(a) is the risk index of worker a, w i is the weight of the i-th influencing factor, v i (a) is the risk scale of the i-th influencing factor of worker a, and S(a) is the safety index of worker a.

[0131] That is, the numerical model expressed by [Equation 1] is a model for calculating the risk index and safety index of worker a to be evaluated, and the risk index is defined as the sum of the products of each of the weights of the plurality of influencing factors and the risk scales multiplied by 10.

[0132] In one embodiment, the plurality of influencing factors may include construction type, basic construction type, site size, construction period, progress rate, month of accident occurrence, time of accident occurrence, work floor, temperature, age, work experience, occupation, and employment type.

[0133] The weight is a predetermined real number between 0 and 1.

[0134] A risk scale for a specific nutritional factor refers to classifying the entire range of data that the nutritional factor may possess into fixed intervals, ranges, types, grades, or classes (hereinafter referred to as "classes"), and assigning numerical values ​​representing the scale to each class of the nutritional factor.

[0135] The weight and risk scale of each of the above-mentioned multiple influencing factors may be determined in advance during the process of constructing the numerical model by the above-mentioned construction module (150), and this will be described later.

[0136] The above-mentioned construction module (150) can construct a numerical model for evaluating the safety index of a construction site worker, and for this purpose, it may include a weight determination module (151), a first risk scale determination module (152), and a second risk scale determination module (153).

[0137] The above weight determination module (151) can determine the weights for each of the plurality of risk factors through the Analytic Network Process (ANP) technique.

[0138] Meanwhile, the above-mentioned multiple influencing factors may be divided into three influencing groups: a construction site group, an environmental group, and a worker group. The above-mentioned multiple influencing factors may include the type of construction, basic work type, site size, construction period, and progress rate belonging to the construction site group; the month of accident occurrence, time of accident occurrence, working floor, and temperature belonging to the environmental group; and the age, work experience, occupation, and employment type belonging to the worker group.

[0139] The above weight determination module (151) is previously " Proposed Worker Safety Index Evaluation Model The weights for the above multiple risk factors can be determined by the method described in Section 3.3 of Part 3, "Calculation of Key Factor Weights."

[0140] FIG. 4 is a flowchart illustrating the process in which the weight determination module (151) determines weights for each of the plurality of risk factors.

[0141] Referring to FIG. 4, the weight determination module (151) can obtain worker survey data for determining the weights of the numerical model (S100).

[0142] The weight determination module (151) can perform pairwise comparisons between the three influence groups, pairwise comparisons between influence factors belonging to the construction site group, pairwise comparisons between influence factors belonging to the environmental group, and pairwise comparisons between influence factors belonging to the worker group based on the worker survey data (S110).

[0143] Afterwards, the weight determination module (151) can generate a super matrix based on the result of the pairwise comparison ((S120), and determine weights for each of the plurality of influence factors based on the generated super matrix (S130).

[0144] In particular, in one embodiment, the weight determination module (151) is previously " Proposed Worker Safety Index Evaluation Model Through the process described in the part, the weights for the above multiple influencing factors can be determined as shown in [Table 1] below.

[0145] Influence group Influencing factors weight Client / Engineering Company Construction company All targets Construction Site Group Types of construction 0.085 0.062 0.071 Basic work 0.050 0.054 0.052 Site scale 0.052 0.057 0.054 Construction period 0.051 0.057 0.054 Completion rate 0.048 0.055 0.052 Environmental group Month of disaster occurrence 0.113 0.119 0.116 Time of disaster occurrence 0.118 0.122 0.120 Working floor 0.099 0.092 0.095 temperature 0.098 0.094 0.096 Worker group years 0.069 0.077 0.072 Work experience 0.080 0.066 0.073 Job type 0.083 0.075 0.079 Employment type 0.053 0.067 0.060

[0146] Meanwhile, the aforementioned multiple influencing factors can be classified into two categories: data-based impact classification and survey-based impact classification. The data-based impact classification includes site size, month of accident occurrence, time of accident occurrence, temperature, age, and work experience, and the survey-based impact classification may include type of construction, basic construction type, construction period, progress rate, work layer, occupation, and employment type.

[0147] The above first risk scale determination module (152) is previously " Proposed Worker Safety Index Evaluation Model Through the process described in Section 4.1 of Part 4.1, "4.1. Derivation of Risk Measures Using Objective Database (DB)," the risk measures for each data-based influencing factor belonging to the data-based influencing classification can be determined based on statistical data obtained from the database (200) of the Korea Occupational Safety and Health Agency (KOSHA).

[0148] FIG. 5 is a flowchart illustrating the process in which the first risk scale determination module (152) determines the risk scale of each data-based influencing factor.

[0149] Referring to FIG. 5, the first risk scale determination module (152) can perform steps S210 to S240 for each of the data-based influencing factors (i.e., site size, month of accident, time of accident, temperature, age, and work experience) (S200).

[0150] In step S210, the first risk scale determination module (152) can generate a frequency distribution table of the data-based influencing factors and the frequency of disaster occurrence based on the statistical data.

[0151] In step S220, the first risk scale determination module (152) can sort the classes of the frequency distribution table in ascending order according to frequency and calculate the log value of the cumulative occurrence proportion.

[0152] In step S230, the first risk scale determination module (152) can perform exponential regression analysis to calculate coefficients of a utility function in the form of a natural logarithm and derive a utility function defined by the calculated coefficients.

[0153] In step S240, the first risk scale determination module (152) can determine the risk scale of the data-based influencing factor based on the derived utility function.

[0154] The above second risk scale determination module (153) is previously " Proposed Worker Safety Index Evaluation Model Through the process described in Section 4.2 of Part 4, "Derivation of Survey-Based Single-Attribute Utility Function," the risk scale of each survey-based influencing factor belonging to the above survey-based influencing classification can be determined based on expert survey data.

[0155] In particular, the first risk scale determination module (152) and the second risk scale determination module (153) mentioned above are " Proposed Worker Safety Index Evaluation Model Through the process described in the part, the risk scale for each of the above multiple influencing factors can be determined as shown in [Table 2] and [Table 3] below.

[0156] Risk scale Month of disaster occurrence Time of disaster occurrence temperature years Site scale Length of service 10 points May, August 10 to 12 o'clock 22~26 degrees 51~55 years old 1~5 people Less than 1 month 9 points June, July, October 12:00–16:00 27~31 degrees 56~60 years old 50~99 people 1~2 months 8 points April, September 08~10 o'clock 17~21 degrees 46~50 years old 10~29 people 2~3 months 7 points March, November 16:00–18:00 12~16 degrees 61~65 years old 5~9 people 6 months to 1 year 6 points january 12~14 o'clock 7~11 degrees 41~45 years old 30~49 people 3~4 months 5 points february 06:00–08:00, 18:00–20:00 2~6 degrees 66 years and older 100~299 people 4~5 months, 1~2 years 4 points 　 0~2:20~22:00 -3 to 1 degree 36~40 years old 300~499 people 5~6 months 3 points 02~04 o'clock 32~36 degrees 31~35 years old 500~999 people more than 2 years 2 points 22~24 hours -13 to -4 degrees 20 years old and under 1,000 or more people 　 1 point 04~06 o'clock -14 degrees or lower, 37 degrees or higher 21~25 years old

[0157] Risk scale Types of construction Basic work Construction period Completion rate Working floor Job type Employment type 10 points Railways, airports, and dams Demolition and renovation work 12 months or younger 60%~80% 30 floors or more scaffolder irregular workers 9 points Factories · Power generation facilities / Roads · Ports · Bridges Steel and timber structure work 12~36 months 20%~60% / 80%~100% 20th to 29th floors Temporary work / Roofing 　 8 points Hospitals / School Facilities / Single-family / Multi-family Housing / Type 1 / Type 2 Neighborhood Living Facilities General and common / temporary construction work / On-site casting and precast concrete earthwork / Earthwork 36 months or older 0%~20% 9th to 19th floors Rebar worker / welder 7 points 　 Surface covering worker / Window and stair worker / Plumbing worker / Electrician and telecommunications worker 　 　 2nd to 9th floors General laborer / Service / Stonemason / Carpenter 6 points Mason / Building Fabric Worker / Waterproofer 1st floor and below Finisher / Plumbing 5 points 　 　 waterproofing work Regular employee 4 points 　 　 3 points 2 points 1 point

[0158] Referring again to FIG. 3, the evaluation data acquisition module (120) can acquire evaluation data for each of the plurality of influencing factors of the worker being evaluated. The evaluation data acquisition module (120) may receive evaluation data from the worker being evaluated, or it may receive evaluation data from a predetermined external device.

[0159] In one embodiment, the evaluation data acquisition module (120) can acquire the type of construction work, basic work type, site size, construction period, progress rate, work month, work time, work floor, and temperature of the construction work being performed by the worker being evaluated, as well as the age, work experience, occupation, and employment type of the worker being evaluated.

[0160] Meanwhile, the above-mentioned risk scale judgment module (130) can determine the risk scale for each of the plurality of influencing factors of the above-mentioned worker based on the evaluation data for each of the plurality of influencing factors of the above-mentioned worker.

[0161] The above safety index calculation module (140) can calculate the safety index of the worker being evaluated by applying the risk scale for each of the plurality of influencing factors of the worker being evaluated to the numerical model described above.

[0162] Meanwhile, the output module (160) can output the calculated safety index of the worker subject to evaluation to an external device. For example, the output module (160) can output the safety index of the worker subject to evaluation through an external output device such as a display device, or output (transmit) the safety index of the worker subject to evaluation to an external device connected via a network. Alternatively, the output module (160) may send a notification to an administrator terminal, etc., when the safety index of the worker subject to evaluation satisfies a predetermined safety standard or does not satisfy it.

[0163] FIG. 6 is a flowchart illustrating a method for evaluating the safety index of construction site workers according to one embodiment of the present invention.

[0164] Referring to FIG. 6, the safety index evaluation system (100) can construct a numerical model for evaluating the safety index of a construction site worker (S300).

[0165] The above safety index evaluation system (100) can obtain evaluation data for each of the above multiple influencing factors of the worker being evaluated (S310).

[0166] Subsequently, the safety index evaluation system (100) can determine the risk scale for each of the plurality of influencing factors of the worker being evaluated based on the evaluation data for each of the plurality of influencing factors of the worker being evaluated (S320), and can calculate the safety index of the worker being evaluated by applying the risk scale for each of the plurality of influencing factors of the worker being evaluated to the numerical model (S330).

[0167] Meanwhile, according to an embodiment, the safety index evaluation system may include at least one processor and memory that stores a program executed by said processor. The processor may include a single-core CPU or a multi-core CPU. The memory may include high-speed random access memory and may include one or more non-volatile memories, such as magnetic disk storage devices, flash memory devices, or other non-volatile solid-state memory devices. Access to memory by the processor and other components may be controlled by a memory controller.

[0168] Meanwhile, the method for evaluating the safety index of a construction site worker according to an embodiment of the present invention may be implemented in the form of computer-readable program instructions and stored on a computer-readable recording medium. A computer-readable recording medium includes all types of recording devices in which data that can be read by a computer system is stored.

[0169] The program instructions recorded on the recording medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of software.

[0170] Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Additionally, computer-readable recording media may be distributed across networked computer systems, allowing computer-readable code to be stored and executed in a distributed manner.

[0171] Examples of program instructions include not only machine code, such as that generated by a compiler, but also high-level language code that can be executed by a device that processes information electronically using an interpreter, such as a computer.

[0172] The hardware device described above may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa.

[0173] The foregoing description of the present invention is for illustrative purposes only, and those skilled in the art will understand that other specific forms can be easily modified without altering the technical spirit or essential features of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive. For example, each component described as a single unit may be implemented in a distributed manner, and components described as distributed may likewise be implemented in a combined form.

[0174] The scope of the present invention is defined by the claims set forth below rather than by the detailed description above, and all modifications or variations derived from the meaning and scope of the claims and equivalent concepts thereof should be interpreted as being included within the scope of the present invention.

Claims

Claim 1 A method for evaluating the safety index of a construction site worker, comprising: a numerical model construction step for constructing a numerical model to evaluate the safety index of an individual worker at a construction site—wherein the safety index is determined by a plurality of predefined influencing factors, and the numerical model is defined as shown in [Formula 1] below; an evaluation data acquisition step for the safety index evaluation system to acquire evaluation data for each of the plurality of influencing factors of the individual worker to be evaluated; a risk scale determination step for the safety index evaluation system to determine a risk scale for each of the plurality of influencing factors of the individual worker to be evaluated based on the evaluation data for each of the plurality of influencing factors of the individual worker to be evaluated; and a safety index calculation step for the safety index evaluation system to calculate the safety index of the individual worker to be evaluated by applying the risk scale for each of the plurality of influencing factors of the individual worker to be evaluated to the numerical model. [Formula 1] (Here, n is the number of the aforementioned multiple influencing factors, V(a) is the risk index of worker a, w i is the weight of the i-th influencing factor, v i (a) is the risk scale of the i-th influencing factor of worker a, and S(a) is the safety index of worker a) Claim 2 In claim 1, the numerical model construction step further includes a weight determination step for determining weights for each of the plurality of influence factors through an Analytic Network Process (ANP) technique, wherein the plurality of influence factors are classified into three influence groups: a construction site group, an environmental group, and a worker group, and the plurality of influence factors include: construction type, basic work type, site scale, construction period, and progress rate belonging to the construction site group; month of accident occurrence, time of accident occurrence, working layer, and temperature belonging to the environmental group; and age, work experience, occupation, and employment type belonging to the worker group, and the weight determination step comprises: a step of acquiring worker survey data for determining weights of the numerical model; and a step of performing pairwise comparisons between the three influence groups, pairwise comparisons between influence factors belonging to the construction site group, pairwise comparisons between influence factors belonging to the environmental group, and pairwise comparisons between influence factors belonging to the worker group based on the worker survey data. A method for evaluating the safety index of a construction site worker, comprising the step of generating a super matrix based on the results of the pairwise comparison and determining weights for each of the plurality of influencing factors based on the generated super matrix. Claim 3 In claim 1, the plurality of influence factors are classified into two categories: data-based influence classification and survey-based influence classification, and the plurality of influence factors include site size, month of accident occurrence, time of accident occurrence, temperature, age, and work experience belonging to the data-based influence classification; and construction type, basic construction type, construction period, progress rate, work layer, occupation, and employment type belonging to the survey-based influence classification, and the numerical model construction step comprises a first risk scale determination step for determining the risk scale of each data-based influence factor belonging to the data-based influence classification based on statistical data obtained from the database of the Korea Occupational Safety and Health Agency (KOSHA); A method for evaluating the safety index of a construction site worker, comprising a second risk scale determination step for determining the risk scale of each survey-based influencing factor belonging to the survey-based influencing classification based on expert survey data, wherein the first risk scale determination step comprises: for each of the data-based influencing factors, a step of generating a frequency distribution table of the data-based influencing factor and the frequency of accident occurrence based on the statistical data; a step of sorting the classes of the frequency distribution table in ascending order according to frequency and calculating the log value of the cumulative occurrence proportion; a step of performing exponential regression analysis to calculate the coefficients of a utility function in the form of a natural logarithm; a step of deriving a utility function defined by the calculated coefficients; and a step of determining the risk scale of the data-based influencing factor based on the derived utility function. Claim 4 A method for evaluating the safety index of construction site workers according to claim 1, wherein the plurality of influencing factors include type of construction, basic type of work, site size, construction period, progress rate, month of accident occurrence, time of accident occurrence, work floor, temperature, age, work experience, occupation, and employment type. Claim 5 In Paragraph 4, the weights for each of the above-mentioned plurality of influencing factors are a method for evaluating the safety index of construction site workers defined as shown in [Table 1] below. [Table 1] Claim 6 In Paragraph 4, the risk scale for each of the plurality of influencing factors is a method for evaluating the safety index of construction site workers as defined in [Table 2] and [Table 3] below. [Table 2] [Table 3] Claim 7 A computer-readable recording medium having a computer program that performs the method described in any one of paragraphs 1 through 6. Claim 8 A computer program stored on a computer-readable recording medium that performs the method described in any one of paragraphs 1 through 6. Claim 9 A safety index evaluation system for a construction site worker comprising at least one processor; and a memory in which a computer program is stored, wherein, when the computer program is executed by the at least one processor, the safety index evaluation system performs a method described in any one of claims 1 to 3. Claim 10 A system for evaluating the safety index of a construction site worker, comprising: a storage module storing a numerical model for evaluating the safety index of an individual worker at a construction site—wherein the safety index is determined by a plurality of predefined influencing factors, and the numerical model is defined as shown in [Formula 1] below; an evaluation data acquisition module for acquiring evaluation data for each of the plurality of influencing factors of the individual worker to be evaluated; a risk scale determination module for determining the risk scale for each of the plurality of influencing factors of the individual worker to be evaluated based on the evaluation data for each of the plurality of influencing factors of the individual worker to be evaluated; and a safety index calculation module for calculating the safety index of the individual worker to be evaluated by applying the risk scale for each of the plurality of influencing factors of the individual worker to be evaluated to the numerical model. [Formula 1] (Here, n is the number of the aforementioned multiple influencing factors, V(a) is the risk index of worker a, w i is the weight of the i-th influencing factor, v i (a) is the risk scale of the i-th influencing factor of worker a, and S(a) is the safety index of worker a) Claim 11 In Clause 10, the safety index evaluation system for construction site workers further includes a weight determination module that determines weights for each of the plurality of influencing factors through the Analytic Network Process (ANP) technique, wherein the plurality of influencing factors are divided into three influence groups: a construction site group, an environmental group, and a worker group, and the plurality of influencing factors include: the type of construction, basic work type, site scale, construction period, and progress rate belonging to the construction site group; and the month of accident occurrence, time of accident occurrence, working floor, and temperature belonging to the environmental group; A safety index evaluation system for construction site workers that includes age, work experience, occupation, and employment type belonging to the above worker group, wherein the weight determination module acquires worker survey data for determining weights of the above numerical model, performs pairwise comparisons between the three influence groups based on the worker survey data, pairwise comparisons between influence factors belonging to the above construction site group, pairwise comparisons between influence factors belonging to the above environmental group, and pairwise comparisons between influence factors belonging to the above worker group, generates a super matrix based on the results of the pairwise comparisons, and determines weights for each of the plurality of influence factors based on the generated super matrix. Claim 12 In Clause 10, the plurality of influence factors are classified into two categories: data-based influence classification and survey-based influence classification, and the plurality of influence factors include site size, month of accident occurrence, time of accident occurrence, temperature, age, and work experience belonging to the data-based influence classification; and construction type, basic construction type, construction period, progress rate, work layer, occupation, and employment type belonging to the survey-based influence classification, and the safety index evaluation system for construction site workers includes a first risk scale determination module that determines the risk scale of each data-based influence factor belonging to the data-based influence classification based on statistical data obtained from the database of the Korea Occupational Safety and Health Agency (KOSHA); A safety index evaluation system for construction site workers that includes a second risk scale determination module for determining the risk scale of each survey-based influencing factor belonging to the survey-based influencing classification based on expert survey data, wherein the first risk scale determination module, for each of the data-based influencing factors, generates a frequency distribution table of the data-based influencing factor and the frequency of accident occurrence based on the statistical data, sorts the classes of the frequency distribution table in ascending order according to frequency, calculates the logarithm of the cumulative occurrence proportion, performs exponential regression analysis to calculate the coefficients of a utility function in the form of a natural logarithm, derives a utility function defined by the calculated coefficients, and determines the risk scale of the data-based influencing factor based on the derived utility function. Claim 13 In Clause 10, the above-mentioned multiple influencing factors include a safety index evaluation system for construction site workers, comprising type of construction, basic type of work, site size, construction period, progress rate, month of accident occurrence, time of accident occurrence, work floor, temperature, age, work experience, occupation, and employment type. Claim 14 In Clause 13, the weights for each of the aforementioned multiple influencing factors are defined as a safety index evaluation system for construction site workers as shown in [Table 1] below. [Table 1] Claim 15 In Clause 13, the risk scale for each of the above-mentioned multiple influencing factors is a safety index evaluation system for construction site workers defined as shown in [Table 2] and [Table 3] below. [Table 2] [Table 3]

Images