Integrating Human Performance Factors to Improve Occupational Risk Assessment

In recent years, occupational health and safety (OHS) has become a public health priority and a major concern in the context of human resources management, particularly in high-risk industries. This is reflected simultaneously in a tightened regulatory framework and growing market pressure for companies to be aware of the need to control their risks. Let's remember that risks are events that prevent a company from achieving its strategic objectives [1, 2]. From now on, risk management must be a commercial logical [3]. Consequently, occupational risk assessment is a crucial step in any prevention approach [4, 5]. Implementing a prevention approach will help to improve the company's step-by-step human and economic performance. To this end, occupational risk assessment (ORA) is the starting point for guaranteeing the sustainability of any business, and above all for avoiding any physical, material or environmental damage type [6]. It is an essential element in the development of a company's risk prevention and management policy [7, 8].

Several approaches have been developed for the assessment of occupational risks, using different methods. However, few works have addressed the integration of human operator performance in this assessment although most occupational accidents are directly or indirectly caused by human operators. In this paper, we propose a hybrid model of occupational risk assessment by integrating human operator performance factors as a key component of the occupational health and safety management programme. The aim of this work is twofold:

The remainder of this document is organized as follows: The next section presents literature review on the different approaches of occupational risk assessment that have already been developed. While section 3 presents the proposed approach to improving conventional ORA. Section 4 is reserved for illustrating the proposed approach through its application to a cement manufacturing process. Application results are discussed in section 5, and the final section offers a conclusion.

Despite its use in systems engineering, SADT can also be adapted to other fields, such as OH&S risk management through ORA. It should also be noted that, although SADT and ORA are distinct concepts, they can be complementary in risk management. By combining the two concepts, them complementarily in risk and occupational hazard management may be highlighted. Indeed, SADT can be a very useful tool for visualizing and analyzing systems, and in so doing, it can help to improve traditional ORA. This SADT and ORA combination offers a more comprehensive and effective approach to Occupational Risk Assessment, with a focus on Human Operator Factors integration, hence Occupational Risk and Human Performance Assessment new name or ORA-P_HO.

3.1 SADT

SADT was developed by Ross [43] following work (1969-1973) on problem solving dating back to 1950s at Softech. SADT, or Structured Analysis and Design Technique, is a graphical systems modeling method for representing and analyzing complex systems. It is a graphical language widely used to describe complex systems in communicative design, military planning and computer-aided manufacturing [44].

SADT aim is to enable experts to view problems from different points of view and levels of abstraction. SADT diagrams used enable a system to be broken down into subsystems and relationships between them to be represented. SADT can be applied to a variety of fields, including systems engineering and process management.

It should also be noted that there isn't a great deal of research into SADT use in OHS management, and in particular its integration into an ORA approach. However, some studies indicate that SADT is necessary formalism for a number of uses [45]. For example, Santarek and Buseif [46] propose SADT use to create high-level system design specifications for flexible manufacturing systems. Cogram and Epelman [47] explain a procedure for describing a service operation (individual tax return) using SADT diagrams at each stage. Kim et al. [48] discuss IDEF models’ integration to bring richer semantics to business modeling. Whitman et al. [49] describe a procedure for converting static models into simulation models using certain model rules and annotations. Other work has been deployed to perform functional and informational modeling of production [50], other SADT work has been used to solve technical production problems in vibratory-centrifugal reinforcement to obtain the required part quality parameters [51].

As part of this work, is to propose integration of SADT into EvRP approach. Indeed, structural and functional modelling of a system is an essential step in any risk analysis. The objective of this step is to identify, describe and prioritize all activities that must be accomplished in a complex industrial process. It provides a simple graphical representation, in diagrams form called actigrams formed of boxes and arrows where boxes represent activities and arrows data processed (information or objects) by activities. In the case, for example, tanning process, inputs are raw hides, chemicals, water, energy, mechanisms or resources that enable the functions of the process to be performed are leather preparation machines, tanning barrels, pumps, agitators and drying equipment. The constraints, which generally represent physical conditions for carrying out functions, are temperatures (tanning baths, heat treatment, drying), pH, chemicals concentration products etc. The outputs are tanned leather, residues or wastes (solid, liquid and gas). All information collected by this formalism and related to different functions of the process can constitute danger sources and generate risks and occupational diseases. These effects include chemical risks due to massive use of dangerous substances (skin irritation and dermatitis, chronic respiratory problems, allergies…), physical risks related to machines, especially rotating (amputations, injuries, falls,), musculoskeletal disorders due to uncomfortable or repetitive postures and possibly biological risks (bacterial infections) when handling animal material. All this information will be used to conduct an ORA.

3.2 ORA

ORA is a regulatory process designed to identify, assess and prevent OHS-related risks. The aim is to implement appropriate preventive measures to ensure OHS management. It is a mandatory process designed to identify hazards that employees are likely to encounter in workplace, and associated risk factors. These risks are classified according to their severity, probability of occurrence and potential number of employees affected. Once a risk assessment has been carried out, results must be transcribed into a single document for consistent sake, convenience and legal traceability [52].

3.3 Human operator performance as a precursor to OHS performance

OHS management is described as science and art of anticipating, recognizing, assessing and controlling occupational hazards in workplace [53]. Some research shows that determining OHS performance history plays a major role in effectively reducing number of workplace accidents [54, 55]. It should be remembered that, in addition to factors such as the work environment and management practices, which have been taken into account as constituting elements of OHS performance antecedents, the same authors Neal and Griffin [56] assert that Human factors have also been noted as OHS performance antecedents. As a result, Human performance plays a significant role in the process of improving OHS performance. Reiman and Pietikäinen [57] assert that "safety indicators can play a key role in providing information and knowledge on organizational performance and thus on OHS performance, motivating employees to work on safety while mobilizing their skills and knowledge and thus increasing organizational safety potential". In conclusion, the performance of the Human Operator depends on three factors: motivation, knowledge and skills. For all these reasons, we decided to take a closer look at the importance of human operator performance in occupational risk assessment. In addition, human performance depends on a number of essential factors, which we will detail in next section.

3.4 Occupational risk and human performance assessment ORA-P_HO

In this article, an Improved ORA is proposed integrating human operator performance for occupational risks assessment linked to production systems. In addition to occupational risk parameters, this new approach considers factors that affect human operator performance.

The first stage of the proposed approach consists in carrying out a technical (structural) and functional analysis of production system to be studied, by modeling it using SADT method. Results of this first step are system's different activities identification, their interactions and the information flows within and between different levels of the system in question. The second stage concerns ORA implementation. Input data for this step, such as activity, hazard sources, risks and effects, are provided using SADT formalism.

Risk is then assessed in terms of injury severity (S), occurrence probability (P), rate exposure (f) and protection level (PL). These parameters characterization according to four-level ordinal scales is illustrated in Table 1.

Table 1. Occupational risk assessment parameters

By multiplying these parameters, real risk may be assessed, taking into account preventive measures already in place:

$R=(f \times E \times S) / P L$           (1)

Risk is represented by an ordinal scale ranging from class 1 (negligible risk) to class 4 (very high risk), as shown in Table 2.

Table 2. Risk scale

As employees are the first to be affected by occupational hazards to which they are exposed, it is essential to integrate their performance into ORA, and this is the third stage in the proposed approach. In the context of this approach, human operator performance can be expressed as:

$\mathrm{P}_{\mathrm{HO}}=M \times C \times K m$           (2)

where,

P_HO: Human operator performance

M: Motivation

Km: Knowledge

C: Competence

These different parameters are evaluated using three-level ordinal scales, as shown in Tables 3 and 4. Indeed, a Safety Management System (SMS), to meet OHS objectives, comprises policies set and practices aimed at positively involving workers' behaviors in facing's risks, with the aim of reducing their dangerous acts. Exposure, on the other hand, refers to duration and resources number that can lead to hazards for people [58]. Risk exposure has also been shown to influence organizational OHS participation [11].

ORA is the commonly used method for assessing activities risk exposure performed by workers.

As the importance of workers' personal knowledge, skills and motivations in OHS management can be explained by their physical proximity to sources of danger. These factors can be integrated into occupational in risk assessment phase; hence, the new name Occupational Risk and human operator performance Assessment, or ORA-P_HO. This combination is justified by the fact that implicit assumptions in activity-based risk assessment suggest that these resource-based risk assessments, such as knowledge held by Human Operator, can be used to compare activities for future operations. Dejoy et al. [39] state that such factors as operator behavior plays a key role in OHS issues. OHS-related knowledge thus corresponds to actions linked to health importance and safety that are shared by a people group.

Three-level ordinal scales as shown in Tables 3 and 4 evaluate these different parameters.

Table 3. Human operator performance evaluation parameters

Table 4. Human operator performance scale

The steps in the general procedure of the proposed approach are shown in Figure 1.

1.png

Figure 1. Flow chart of the proposed approach

Workers at SCIMAT cement plant are exposed to numerous occupational hazards, which have been listed in the table (improved ORA results). Assessment of these risks in terms of probability, exposure, severity and availability of protective means, show that among risks identified, chemical, ergonomic and traffic risks are the highest risks to which workers are exposed during various cement production activities.

In fact, the raw meal preparation, cement production and packaging/shipping phases are the most critical where human intervention is indispensable, working conditions and means of prevention are inadequate or non-existent.

The results also reveal that workers are exposed to moderate risks, such as the physical environment (noise, thermal discomfort, lighting and dust inhalation) and the risks of falls from ground level and heights caused by poor working conditions.

Furthermore, the results obtained show that the performance of human operators within the cement plant varies from low to high. In fact, in the raw meal preparation phase, the performance of human operators is low, as the majority of workers are unmotivated; lack the necessary knowledge and skills, and the majority are not permanent workers.

However, in the cement production, packaging and shipping phases, the performance of human operators is average, requiring them to improve their skills and knowledge through training and improved working conditions, in order to motivate them further.

On the other hand, human operator performance is high in the flour-baking phase, where human operator tasks (intervention) are limited to inspection and supervision of the indicated phase.

In order to draw more comprehensive conclusions and explore other dimensions of results that are not directly apparent in Table 6, an in-depth analysis will be realised.

Table 6. Data required for correlation analysis

Table 6 highlights several parameters influencing risk overall score R and P_OH for each identified activity. This allowed us to construct an enriched table, including several variables in order to explore in depth relationships between these variables. This data exploration aims to highlight specific relationships or trends, in particular by identifying whether certain variables significantly influence R and P_OH, which would help to validate and confirm results presented in Table 6.

For this purpose, a correlation analysis between variables and to see how different variables interact is carried out. This helps to identify which variables strongly influence $R$ and $P_{O H}$ and provide assistance for working conditions improvement in terms of OSH and recommend actions to reduce risks in well-targeted ways.

It serves also indirectly to validate whether the expected logical relationships are confirmed in our data.

Correlation analysis can therefore:

-Identify key relationships between variables $\left(f, E, G, N, R, K_m, C, M, P_{O H}\right)$.

-Use of correlation matrix to highlight significant relationships.

To calculate correlation coefficients between variables $\left(f, E, G, N, R, K_m, C, M, P_{O H}\right)$, data the results provided can be used.

Table 6 is constructed that groups together data needed for correlation analysis. Each row represents an activity as well as the risks associated with it, while columns detail the different parameters allowing risk estimation R and performance P_OH.

This organized structure aims to provide a clear and systematic view for data analysis.

Pearson correlation coefficient is calculated using the following formula:

$r(x, y)=\frac{\operatorname{Cov}(x, y)}{\sigma_x \sigma_y}$           (3)

where,

$\operatorname{cov}(x, y)$: is the covariance.

$\sigma_x$: is the standard deviation of $x$.

$\sigma_y$: is the standard deviation of $y$.

Result gives the correlation matrix of the form given in Table 7.

Table 7. Correlation matrix

The matrix shows relationships between different variables, expressed by Pearson correlation coefficient. Values close to +1 indicate a strong positive correlation, while those close to -1 show a strong negative correlation. Values close to 0 suggest an absence of a significant relationship. A graphical visualisation (Figure 4) is also provided to aid interpretation of key correlations, highlighting significant relationships that could influence the analysis results.

4.png

Figure 4. Correlogram illustrating the correlations between variables

A positive linear relationship exists between variables frequency of occurrence (F) and risk (R), as well as between exposure rate (E) and risk (R). The Pearson correlation coefficients for these pairs are 0.47 and 0.73, respectively. These values indicate a moderate to strong positive relationship between the variables, suggesting that frequent events and increased exposure are generally associated with a higher risk level.

A negative linear relationship is observed between level of protection (N) and the risk (R) with a Pearson correlation coefficient of -0.76. This value indicates a strong and negative relationship, meaning that as the level of protection increases, the risk decreases, which highlights the importance of protective measures to mitigate risks.

Regarding performance parameters a positive linear relationship is also present between competence (C) and human performance P_OH, as well as between motivation and human performance P_OH. The correlation coefficients for these pairs are 0.86 and 0.68, respectively. These values indicate a strong relationship between these variables, highlighting that high skills and increased motivation are key factors in improving human operator performance.

Results of this analysis provide valuable guidance for prioritizing actions based on identified correlations. Prioritization is based on correlation coefficient: stronger correlation, the more factor concerned becomes a priority for action.

Thus, two main fronts emerge. On one hand, it is essential to reduce risks through improved management of frequency, exposure and protection levels. On other hand, improving human performance requires investments in skills, motivation and knowledge of operators. By acting in a targeted manner on these two axes, we can optimize both safety and performance of human operators.

Furthermore, the proposed approach can have different impacts on working conditions improvement:

-The involvement of workers in OHS programs development allows them to benefit from their practical expertise, which facilitates identification, control and management of risks and thus creates a positive dynamic of prevention within company.

-The adoption of negotiated and tested solutions allows a reduction in work accidents and work stoppages, workers empowerment and an increase in productivity.