Hybrid multi-criteria decision making approach proposal for evaluating workloads of aircraft maintenance technicians

Abstract

BACKGROUND:

Aircraft maintenance technicians (AMTs) have the most difficulty in terms of total workload criteria working in line maintenance. This is a very important problem for the Aircraft Maintenance Organization. A systematic and scientific approach is required for its solution.

OBJECTIVE:

This study proposes a new Multi-Criteria Decision Making (MCDM) based approach to evaluating the total workloads of AMTs to identify the most challenging AMT tasks in the aircraft maintenance organization.

METHODS:

A new hybrid MCDM approach is proposed by integrating the Decision-Making Trial and Evaluation Laboratory (DEMATEL) and Technique for Order Preference by Similarity (TOPSIS) methods to compare AMTs on the basis of workloads according to license categories. The hybrid method proposed in this study evaluates the total workload under three main titles: mental, physical, and environmental workload.

RESULTS:

Focusing on AMTs working in line maintenance of an aircraft maintenance organization, this study revealed that the most important workload criteria determined by the DEMATEL method are lower back strain, upper back strain, time pressure, and air temperature criteria. The results of the TOPSIS method showed that the license categories of AMTs are sorted according to the workloads as follows: A, B2, B1, and B1 + B2. The AMTs holding a “Category A” license have fewer workloads than the other categories.

CONCLUSION:

The findings of the study reveal some measures that might allow authorities to minimize the workload of AMTs. In addition, the study contributes to the literature because there are few studies that systematically analyze total workloads by using MCDM methods.

Keywords

Aircraft maintenance technician workload ergonomic MCDM DEMATEL TOPSIS

1 Introduction

All kinds of pressures unfavorably affecting work performance and productivity due to the nature of the work and working environments are evaluated as workload [1]. Physical and mental workloads affect an employee’s health, performance, and productivity. Environmental factors as well as mental and physical ones also considerably affect the total workload level [2, 3].

Although the share of the air transport sector tends to decrease from time to time due to unprecedented crises, people increasingly demand air transport because it is both safer and faster than other means of transport. According to the data published by International Civil Aviation Organization (ICAO), airlines worldwide carried nearly 4.5 billion passengers in 38.3 million scheduled flights in 2019. Due to the pandemic that broke out in 2020, the number of passengers decreased by 60% worldwide [4]. Safety and security are among the most important factors that are taken into consideration in the aviation sector due to the high frequency of flights and increasing demand for air transport. Aircraft maintenance technicians (AMTs) ensure a safe flight by doing their best to fulfill all planned or unplanned maintenance tasks in an aircraft despite the presence of physical and mental difficulties [5 –8]. Aviation maintenance and inspection are complex tasks in which individuals perform various duties under time pressure, minimal feedback, and difficult environmental conditions [9]. AMTs work on hangars or aprons. A hangar is a generally large building used for the maintenance and repair of aircraft at the airport or for storage. The Apron is a defined area at an airport for passenger, mail, and cargo loading-unloading, refueling, maintenance, and parking purposes. While performing the modification, troubleshooting, and part replacement duties in maintenance hangars or aprons, AMTs endure intense physical workloads such as lifting heavy objects and working under improper working conditions as well as mental workloads such as time pressure and stress due to unplanned maintenance and environmental workloads such as heat, noise, lighting, and vibration. Such challenging conditions are known to increase workloads and, in turn, negatively affect the workforce [5 , 11]. It is crucial to reduce the workloads of AMTs to minimum levels in order to ensure flight safety. Evaluation of total workload might be tackled as a Multi-Criteria Decision Making (MCDM) problem encountered in relation to mental, physical, and environmental workload criteria.

The present study proposed a new hybrid approach based on MCDM. In this approach, the criteria weights were determined employing the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method –which is an MCDM method, and AMTs holding different maintenance license categories in line maintenance (A, B1, B2, B1 + B2) were compared by using these criteria weights in Technique for Order Preference by Similarity (TOPSIS) method. The DEMATEL method which originated from the Geneva Research Centre of the Battelle Memorial Institute [12], has long been used widely to reveal the hierarchical structure of criteria. It is used not only to confirm the interactive relationships among various criteria but also to seek the most accurate influential criteria weights. This method can allow researchers to identify whether a particular criterion is affected by other criteria or affects other existing criteria. The literature review revealed that the DEMATEL method was employed in scientific studies for the following purposes: selection of logistics service providers [13], evaluation of ecological safety [14], risk factor analysis in hospitals [15], selection of renewable energy resources [16], criteria interaction management studies [17] and determining workload-related stress levels of air traffic controllers [18], developing an effective security system for airlines [19], determining criteria for selection of aircraft to protect air traffic [20], explaining risky behaviors of pilots [21] and determining physical workload weights of AMTs [22] in the aviation sector. Developed by Hwang and Yoon, the TOPSIS method is based on choosing the closest alternative to the positive ideal solution and the farthest alternative to the negative solution as the best alternative [23]. The positive ideal solution is the combination of all accessible best criteria while the negative ideal solution is composed of all worst criteria attained [24]. According to the literature review, the TOPSIS method was employed in scientific studies for various purposes such as the selection of locations for wind farms [25], the selection of countries for sustainable supply chains [26], the selection of maintenance strategy in the mining industry [27], selection of coating material [28], and forest fire susceptibility mapping [29]. As for the aviation sector, the method was adopted to improve the competence of maintenance staff in an aircraft maintenance organization [30], to select maintenance strategy for aircraft systems [31], to determine the stakeholders’ approaches towards sustainable aviation fuel production [32], and to select the best aircraft for basic training [33]. The difference of the hybrid approach proposed in the present study is that it compares AMTs in terms of total workloads by using DEMATEL and TOPSIS methods according to physical, mental, and environmental workloads criteria. The physical workload criteria used in the present study for the evaluation of total workload are neck, shoulders, elbows, wrists, upper back, lower back, hips, knees, ankles, upper arms, forearms, thighs, the lower leg parts of the body as specified in Cornell Musculoskeletal Discomfort Questionnaire (CMDQ) and Nordic Musculoskeletal Questionnaire (NMQ) [34]. As for the mental workload, the study used time pressure and performance, which are the criteria specified in The National Aeronautics and Space Administration-Task Load Index (NASA-TLX) while environmental workload criteria included vibration, noise, lighting, humidity, and heat. The findings of the study revealed some concerns that might allow authorities to minimize the workload of AMTs. In addition, the study is believed to contribute to the literature because there are few studies that systematically analyze total workloads by using MCDM methods.

Following this introduction, the second section presents information about physical, mental, and environmental workloads, which altogether constitute the total workload. The third section explains the stages followed while implementing the proposed approach. In the fourth section, the application of these methods in the study is explained. Finally, the fifth chapter provides a comprehensive discussion of the findings and concludes with general conclusions and recommendations for future research.

2 Theoretical background

2.1 Physical workload assessment for AMTs

AMTs mostly work in line and base maintenance in aircraft maintenance organizations. Those working in line maintenance work under pressure in the environments such as aprons that do not require the presence of a hangar while base maintenance technicians are responsible for the modification and maintenance of aircraft, troubleshooting, and part replacements in hangars or workshops in accordance with the approved standards [35]. AMTs lift heavy loads and perform high-risk work, which might result in musculoskeletal disorders (MSDs) during maintenance and fixing procedures. MSDs often occur during simple body movements such as bending, standing up, holding, grasping, twisting and stretching, etc. These movements normally do not risk our health in our routine daily life. What makes them harmful is continuous repetitions while working, the obligation to apply extra force, and sudden movements [36]. In this respect, MSDs are traumas that slowly develop in time rather than one-time traumas [37]. Work-related MSDs including the lower back, shoulders, legs and feet, neck, arms, and hands are frequently reported among AMTs in the literature [38 –44].

Physical workload measurement methods include questionnaires and self-reported data, checklists that have limiting values for the assessment of specific physical workloads types such as NMQ and CMDQ, task and posture analysis such as Rapid Upper Limb Assessment (RULA), Plan för Identifiering av Belastningsfaktorer (PLIBEL) and Occupational Repetitive Action Index (OCRA), etc., biomechanical, energy/cardiopulmonary, muscular, psychophysical and epidemiological factors [45 –49].

NMQ is administered to identify health problems that occurred in the last 12 months or 7 days on 9 symptom areas (neck, shoulders, upper back, lower back, wrists-hands, ankles-feet, femur–hips, knees) [50]. CMDQ, on the other hand, evaluates body posture and similar factors in 18 different parts of the body including the neck, shoulder (left-right), upper back, upper arm (left-right), lower back, forearm (left-right), wrist (left-right), hip, upper leg (left-right), knee (left-right) lower leg (left-right). It also measures the frequency and severity of aches as well as their effects on work performance [51 –53].

2.2 Mental workload assessment for AMTs

AMTs often have to deal with intense mental challenges in addition to physical ones. A survey administered to 41 AMTs in Indonesia found that 11 technicians experienced medium-level of mental workload and 30 technicians had to cope with high levels of mental workload [54]. Mental workload is often defined as the disparity between the processing capacity of the human information processing system and the capacity demanded to accomplish a given task or activity. In essence, it can be understood as the level of processing capacity required [55 –59]. Despite the lack of a specific definition for mental workload in International Organization for Standardization (ISO) standards, it is explained under two main titles: mental stress and mental strain [60]. Mental stress refers to the mental effects of external factors on individuals while mental strain is the effect of this stress on individuals.

Mental workload can be assessed through three distinct approaches, namely subjective techniques, physiological techniques, and performance-based techniques. Subjective techniques are widely employed due to their applicability and ease of use, utilizing one-dimensional and multidimensional scales. These techniques involve the application of scales, both one-dimensional and multidimensional, to determine and evaluate mental workload [61]. NASA-TLX (NASA Task Load Index), MCH (Modified Cooper-Harper Scale), and SWAT (Subjective Workload Assessment Technique) are commonly favored methods. However, the NASA-TLX technique possesses greater efficacy in representing mental workload compared to the other two [62, 59]. NASA-TLX is a highly recognized method used while measuring workload due to its wide range of use and more accurate measurements at low workload levels [3]. In addition, the related studies reported that it is more reliable and more widely acknowledged than other similar mental workload measurement methods [63]. NASA-TLX evaluates mental workload according to the following six criteria: mental demand, physical demand, effort, performance, temporal demand, and frustration [58].

2.3 Environmental workload assessment for AMTs

Environmental workload refers to workloads stemming from the existing conditions of a working environment. Among the factors affecting environmental workload are improper heat, noise, lighting, humidity, vibration, harmful chemical gases, and dust in the area. If such negative factors have considerable effects and are repetitive, work performance is negatively affected and work stress increases [49, 64].

AMTs also suffer from environmental workloads stemming from existing working conditions. Especially, tough seasonal conditions such as rain, snow, storm, and environmental stress sources including darkness and loud sound from aircraft engines affect line maintenance technicians since they work outdoors in the apron [5, 11].

It has been observed that the total workload assessment of employees working in various sectors is considered a MCDM problem that arises from mental, physical, and environmental workload criteria. However, no study has been found regarding Aircraft Maintenance technicians’ workload. However, aircraft maintenance technicians are responsible individuals for the maintenance and repair of aircraft. While performing their duties, they are exposed to heavy and repetitive physical and mental loads. Additionally, environmental conditions are another important factor affecting their workload. It is extremely important to investigate AMTs’ workload and reduce it when necessary for safe flight operations. Total workload assessment for AMTs can be considered an MCDM problem that arises from physical, mental, and environmental workload factors. In this study conducted in a line maintenance department, various criteria affecting the workload were identified by obtaining expert opinions.

3 Proposed hybrid multi-criteria decision-making approach to determine multidimensional workload for AMTs

The proposed hybrid approach used the DEMATEL method to evaluate the weights of the criteria and the TOPSIS method to assess the alternatives according to the criteria in this study. Matrix calculations are used in the application stages of DEMATEL and TOPSIS methods. The steps of the proposed hybrid MCDM approach are shown in Fig. 1.

Fig. 1

The steps of the proposed hybrid MCDM approach.

Stage 1. Determining Decision Makers (DMs), workload criteria, and AMT alternatives

First of all, DMs are determined as DM_k ={DM₁, DM₂, …, DM_l}; k = (1,2, ... ,l). The criteria are defined as C_i ={C₁, C₂, …, C_n}; i = (1,2, ... ,n) and alternatives as A_t ={A₁, A₂, …, A_v}; t = (1,2, ... ,v).

Stage 2. Determining the criteria weights by using the DEMATEL method

In this stage, the DEMATEL method was implemented in five steps [65].

Step 2.1. Creating the average direct relation matrix

In the DEMATEL method, the direct relation matrix [X] is created first as shown in Equation 1. The relationships between the criteria are expressed with a direct-relation matrix to obtain the criteria weights. DMs determine the relationships between criteria by using a 0–4 scale as displayed in Table 1.

Table 1

Pairwise comparison scale [66]

Numerical value	Definition
0	No Effect
1	Low Effect
2	Medium Effect
3	High Effect
4	Very High Effect

$X = [\begin{matrix} x_{11} & x_{12} & \dots & x_{1 n} \\ x_{21} & x_{22} & \dots & x_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ x_{nn} & x_{nn} & \dots & x_{nn} \end{matrix}] i, j \in {1, 2, \dots . ., n}$ (1)

In this matrix, x_ij shows the degree of the direct effect of criterion i on another criterion j in order to account for the relation.

The average direct relationship matrix [A] is obtained by calculating the means of [X] that were determined by each expert. [A] matrix displays the common decision of DMs. $A = [\begin{matrix} a_{11} & a_{12} & \dots & a_{1 n} \\ a_{21} & a_{22} & \dots & a_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ a_{nn} & a_{nn} & \dots & a_{nn} \end{matrix}]$ (2)

In this matrix, a_ij indicates the average relationship of criterion i to criteria j.

Step 2.2 Normalizing the direct relation matrix

The normalized direct relation matrix [M] is obtained by using Equations 4 as well as [A], which was obtained in Step 1. $k = min (\frac{1}{max_{i} \sum_{j = 1}^{n} | a_{ij} |}, \frac{1}{max_{j} \sum_{i = 1}^{n} | a_{ij} |})$ (3) $M = k . A$ (4)

Step 2.3 Creating the total relation matrix

In this step, the total relationship matrix [T] is calculated using the normalized direct relationship matrix [M] and Equation 5.

$\begin{matrix} T & = M + M^{2} + M^{3} + \dots \\ = \sum_{i = 1}^{\infty \leftarrow} M^{i} = M {(I - M)}^{- 1} \end{matrix}$ (5)

Here, I is an identity matrix.

Step 2.4 Determining sender and receiver groups

Equations 8 are used to create sender and receiver groups. D_i indicates the sum of ith line and R_i indicates the sum of ith column in the T matrix. D_i - R_i and D_i + R_i values determine to what extent the ith affects and the relationship with other criteria. $T = {[t_{i, j}]}_{nxn} i, j \in {1, 2, \dots . ., n}$ (6) $D_{i} = \sum_{j = 1}^{n} t_{i, j}$ (7) $R_{i} = \sum_{i = 1}^{n} t_{i, j}$ (8)

The ith criterion with a positive D_i - R_i value has a higher impact on the other criteria and is referred to as the sender. The ith criterion with a negative D_i - R_i value is affected by other criteria and this criterion is called the receiver. On the other hand, while the ith criterion with a high D_i + R_i value is more related to other criteria; the ith criterion with a low D_i + R_i value is less related to the others.

Step 2.5. Determining the criteria weights

Criteria weights are determined by using Equations 10. $w_{i} = \sqrt{{[D_{i} + R_{i}]}^{2} + {[D_{i} - R_{i}]}^{2}}$ (9) $W_{i}^{*} = \frac{w_{i}}{\sum_{i = 1}^{n} w_{i}}$ (10)

Here, w_i is the weight value of ith criterion and $W_{i}^{*}$ indicates the relative weight value.

Stage 3. Ranking the alternatives by using the TOPSIS Method

In this stage, the TOPSIS method was applied in six steps [27].

Step 3.1. Creating the decision matrix

This step involves the creation of the decision matrix [R], in which each criterion is evaluated according to an alternative. This matrix is shown in Equation 11.

$\begin{matrix} R = & [\begin{matrix} r_{11} & r_{12} & \dots & r_{1 n} \\ r_{21} & r_{22} & \dots & r_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ r_{v 1} & r_{v 2} & \dots & r_{vn} \end{matrix}] \\ t = 1, 2, \dots, v; i = 1, 2, \dots, n \end{matrix}$ (11)

Here, r_ti is the value of tth alternative for ith criterion.

Step 3.2. Normalizing the decision matrix, which is obtained in step 3.1., is normalized through vector normalization by using Equation 12. The normalized decision matrix [N] is displayed in Equation 13. $n_{ti} = \frac{r_{ti}}{\sqrt{\sum_{t = 1}^{v} r_{ti}^{2}}} t = 1, 2, \dots, v; i = 1, 2, \dots, n$ (12) $N = [\begin{matrix} n_{11} & n_{12} & \dots & n_{1 n} \\ n_{21} & n_{22} & \dots & n_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ n_{v 1} & n_{v 2} & \dots & n_{vn} \end{matrix}]$ (13)

Here, n_ti is the normalized value of tth altenative for ith criterion.

Step 3.3. Calculating the weighted normalized decision matrix

In this step, the weighted normalized decision matrix [V] is obtained by multiplying the [N] matrix with the relative weights found in Stage 2 using the DEMATEL method.

$V = [\begin{matrix} v_{11} & v_{12} & \dots & v_{1 n} \\ v_{21} & v_{22} & \dots & v_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ v_{v 1} & v_{v 2} & \dots & v_{vn} \end{matrix}] = [\begin{matrix} W_{1}^{*} n_{11} & W_{2}^{*} n_{12} & \dots & W_{n}^{*} n_{1 n} \\ W_{1}^{*} n_{21} & W_{2}^{*} n_{22} & \dots & W_{n}^{*} n_{2 n} \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ ⋮ & ⋮ & ⋮ & ⋮ \\ W_{1}^{*} n_{v 1} & W_{2}^{*} n_{v 2} & \dots & W_{n}^{*} n_{vn} \end{matrix}]$ (14)

Here, v_ti is the weighted normalized value of tth alternative for ith criterion.

Step 3.4. Calculating the positive ideal and negative ideal solutions

Two sets of solutions, The Positive Ideal Solution (PIS) and The Negative Ideal Solution (NIS), are obtained from [V] using Equations 16. As for the benefit type, the set consisting of the best values of V is called PIS (A*), and the set consisting of the worst values of V is called NIS (A^–). If there is an evaluation of the cost type, the set A*consists of the smallest of the values in [V], and A^– consists of the largest of the values in [V]. $\begin{matrix} A^{*} = {(max_{t} v_{ti} i \in I), (min_{t} v_{ti} i \in I^{'})}, \\ v_{i}^{*} = {v_{1}^{*}, v_{2}^{*}, \dots, v_{n}^{*}} \end{matrix}$ (15) $\begin{matrix} A^{-} = {(min_{t} v_{ti} i \in I), (max_{t} v_{ti} i \in I^{'})}, \\ v_{i}^{-} = {v_{1}^{-}, v_{2}^{-}, \dots, v_{n}^{-}} \end{matrix}$ (16)

In Equations 16, I represents the benefit and I′ the cost value.

Step 3.5. Calculating the distances of the alternatives to PIS and NIS

Distances of the alternatives to PIS and NIS ( $S_{t}^{*}$ and $S_{t}^{-}$ ) are calculated by using Equations 18. $S_{t}^{*} = \sqrt{\sum_{i = 1}^{n} {(v_{ti} - v_{i}^{*})}^{2}}$ (17) $S_{t}^{-} = \sqrt{\sum_{i = 1}^{n} {(v_{ti} - v_{i}^{-})}^{2}}$ (18)

Step 3.6. Determining the relative closeness to ideal solution and ranking alternatives

Relative closeness to PIS is calculated by using Equation 19 for all options. $C_{t}^{*} = \frac{S_{t}^{-}}{(S_{t}^{-} + S_{t}^{*})} 0 ⩽ C_{t}^{*} ⩽ 1$ (19)

The alternative with the highest $C_{t}^{*}$ value is determined as the best alternative.

4 Application

In this section, AMTs holding four different aircraft maintenance license categories (A, B1, B2, B1 + B2) were compared according to the working load criteria. Total workload criterion weights of AMTs operating in line maintenance were determined using the DEMATEL method. AMTs were sorted in terms of total workload according to license categories by employing TOPSIS method.

Stage 1. Determining DMs, workload criteria, and AMT alternatives

Firstly, the DMs were selected since they were supposed to evaluate the criteria and alternatives required for the purposes of the method. The complexity of the decision problem will be increased as the dimensions of the investigated problem space get larger; where the alternative and DM profiles amount expand the decision area identified in the problem space and the criteria identified to be considered restrict this area as well as support DMs to reach an effective solution point even an optimal solution might not be possible to be yielded [67]. The existing literature indicates that the DM profile amounts have a decisive impact on the complexity level of the decision problem; in case fewer DM profiles were defined than they should be then the results may be more insensitive to reflect the real-life conditions since all perspectives might not be taken into account, while defining too many profiles unnecessarily enlarges the problem size and leads the problem to insolvency, besides increasing the processing load, causing misuse of resources and making the results more error-prone. Ordinarily, the selection and sequencing problems handled seperately, chiefly due to computer technology and solution algorithms’ limitations. According to enable decision makers to block the additional complexity challenges that would emerge while analyzing large-scaled problems solution space should be identified to obstruct the unnecessary parameters. As the importance and influence on the problem complexity of the problem dimensions are indicated with the existing works of current literature, the determination of the size of the DM set appropriately gains more importance.

In MCDM methods, the determination of the number of DMs (experts) is considered independently of the number of variables. What matters is not the number of DMs but the quality of their expertise and job experience in the subject matter. In fact, increasing the number of DMs can sometimes lead to inconsistencies in the matrices, as experienced and discussed in the literature and in other published articles by the authors. If there are numerous DMs, efforts are made to reduce their opinions (so few DMs) by reaching a consensus. Numerous examples in the literature support the idea of having DMs at these numbers. As valuable instances of the rich and deep group decision-making (GDM)-MCDM literature; the DMT was composed of five to seven DMs in the studies of Adar and Delice [68], Yazgan and Yılmaz [69], Can and Delice [70]; and, in the studies of Delice and Can [71], Karadağ and Delice [72] four DMs were introduced as DMT members; and, Yılmaz Kaya [67], Can and Delice [73] employed three and fewer DMs. In fact, there are many studies in the literature where the number of DMs is not specified [74, 76]. It should be noted that in the determination of the number of DMs to be included, the quality dimension should be taken into account but not the quantities, and, that the number of DMs deemed appropriate due to the definition of the problem addressed has to be assigned in the decision process.

The DMs of the present study were four employees who work as certifying staff (CS) in line maintenance and have an average of a 10-year of experience. Certifying personnel is authorized by a Part-145 approved maintenance organization to certify the completion of maintenance in accordance with Part-145 [77]. DMs have been selected with professional experience or knowledge in aircraft maintenance and they are drawn from the senior AMTs and mechanics and avionics (who have B1 and B2 licenses) at line stations and aircraft maintenance managers in various aircraft maintenance organizations. CSs also receive occupational health and safety training especially on risks of musculoskeletal disorders from experts at the workplace every 2 years and have certificates on this subject due to passing the exams successfully.

Secondly, the workload criteria accessed through the literature review were examined by DMs, and 20 different workload criteria were determined under three categories: physical, environmental, and mental workload criteria. As for the physical workload criteria, the decision makers used strains in body parts specified in NMQ and CMDQ while two dimensions from the NASA-TLX method were determined as mental workload criteria, and the most common environmental factors affecting line maintenance staff were taken as environmental workload criteria within the scope of this study. These 20 workload criteria can be listed as follows:

C₁: neck strain	C₁₁: forearm strain
C₂: shoulder strain	C₁₂: femur strain
C₃: elbow strain	C₁₃: lower leg strain
C₄: wrist strain	C₁₄: strain experienced when working in a very hot or very cold environment
C₅: upper back strain	C₁₅: strain experienced when working in a noisy environment
C₆: lower back strain	C₁₆: strain experienced when working under low and high humidity conditions
C₇: hip strain	C₁₇: strain experienced when working in an inadequately lighted environment
C₈: knee strain	C₁₈: strain experienced when working in a vibrating environment or while using hand tools
C₉: ankle strain	C₁₉: time pressure
C₁₀: upper arm strain	C₂₀: performance level

The final step in this stage was to determine four different aircraft maintenance license categories for the AMTs working in line maintenance [67].

Category A: A technician holding this category aircraft maintenance license is responsible for minor maintenance tasks and simple part placements in accordance with the limitations specified in the authorization certificate. When the required tasks are successfully fulfilled, the technician can issue a Certificate of Release to Service within the limitations specified by the aircraft maintenance organization he works for.

Category B1: A technician holding this category aircraft maintenance license is authorized to carry out maintenance tasks in power units, airframes, and mechanical and electrical systems, to detect faulty parts and fix them, and to perform certain tests to check whether the system works on aviation systems. When the required tasks are successfully fulfilled, the technician can issue a Certificate of Release to Service or take part in these activities as support personnel.

Category B2: This license category authorization allows the holder to carry out maintenance tasks on the avionic and electrical systems. In addition, the technician can perform electrical and avionic operations that require simple tests to detect whether power systems and mechanical systems work properly. When the required tasks are successfully fulfilled, the technician can issue a Certificate of Release to Service or take part in these activities as support personnel.

Category B1 + B2: The technician holding this category license is authorized to carry out all maintenance tasks stated in both category B1 and category B2 licenses.

Stage 2. Determining the criteria weights by using the DEMATEL method

In this stage, workload criteria weights were determined using the DEMATEL method by following the steps explained below.

Step 2.1. Creating the average direct relation matrix

In this step, the DMs firstly developed pairwise comparison matrices by using the comparison scale presented in Table 1. Then, these matrices were combined with arithmetic means and the average direct relation matrix was obtained and displayed in Table 2.

Table 2
Average direct relation matrix [A]

C _i C ₁ C ₂ C ₃ C ₄ C ₅ C ₆ C ₇ C ₈ C ₉ C ₁₀ C ₁₁ C ₁₂ C ₁₃ C ₁₄ C ₁₅ C ₁₆ C ₁₇ C ₁₈ C ₁₉ C ₂₀

C ₁ 0 2 3,25 2,25 2,5 2,25 2,75 1,75 2,75 3 2,25 2,5 2,25 1,5 2 2 2,75 3 1,75 1,75

C ₂ 3,25 0 3 3 1,5 1,5 3 3,5 3 1,75 1,75 2,25 3,25 1,75 1,75 2 2,5 2,25 1,75 1,5

C ₃ 1 1 0 1 1 0,75 1,75 2 1,25 1,5 1,25 1,25 1,25 1,5 1,25 1,25 2 2,25 1,5 1,5

C ₄ 1,5 1,5 2,75 0 1,25 1,5 1,75 2,25 2 1,5 1,25 2,25 1,75 1,25 1,5 1 2,25 2,5 1,25 1,25

C ₅ 4 4 3,75 4 0 3,5 4 3,75 4 4 3,75 4 4 2,5 2,25 2,5 3 3 2,25 2,25

C ₆ 3,75 4 4 4 3,75 0 4 4 4 4 3,75 4 4 2,5 2,25 2,25 3,25 3,25 2,25 2,25

C ₇ 1,5 2 2,25 1,75 1 0,75 0 1,25 1,5 2,25 2,25 1,5 1,25 1,25 1,75 1,25 2,25 2 1,25 1,25

C ₈ 3 1,5 3 2,5 1 1,25 2,75 0 3 3,25 2,75 2 2 1,25 1,5 1,5 2,25 2,25 1,5 1,5

C ₉ 1,25 1,25 1,25 1 0,75 1,25 2,5 2 0 1,75 1,5 1,5 1,5 1,5 1,75 1,25 2,5 2,25 1,25 1,5

C ₁₀ 0,75 1,25 2,5 3 1,25 1 2,75 1,5 2 0 1,5 1,75 2 1,25 1,75 1,5 2,25 2,5 1,5 1,5

C ₁₁ 1,25 1,75 1,5 2,25 1 1 2,75 1,75 1,75 1,75 0 1,75 1,75 1,5 1,75 1,75 2,5 2,25 1,25 1,25

C ₁₂ 1 1 2,5 1 0,5 0,75 2 1 2 1,25 1 0 1 1 1,25 1,5 2 2 1,25 1,25

C ₁₃ 1,25 1,25 2,5 1,25 1,25 1 1,5 1,5 2,5 1,5 1,5 1,5 0 1,75 1,25 1,5 1,75 1,75 1 1

C ₁₄ 3 3 3,75 3 2,25 2,25 3,5 3,5 3,75 3,5 3,25 3,75 3,5 0 3,25 2,5 3,25 3,25 1,75 1,75

C ₁₅ 2,25 2,5 2,5 2,5 1,25 1,25 2,5 2 1,5 1,5 1,5 2,5 2,5 1,25 0 1,25 2,5 2,25 1,25 1,25

C ₁₆ 3 2,75 3 2,75 2 1,75 3 2,5 3 3 3 3 3 1,75 3 0 3 3 1,75 1,75

C ₁₇ 1,5 1,5 2,5 1,5 1,25 1,25 2,5 1,5 1,5 1,5 1,5 1,5 1,5 1 1,5 1,5 0 1,5 1,5 1,5

C ₁₈ 2,25 1,5 2,75 1,5 1,5 1,5 2,75 1,75 1,75 1,75 1,75 1,5 1,75 1,25 1,5 0,75 2,5 0 1,25 1,25

C ₁₉ 3 3,25 3 3 3,25 3 3,5 3 3,25 3,25 3 3,25 3 2,75 2,75 2,5 3,25 3 0 2,5

C ₂₀ 3,5 2,75 2,75 2,5 2,75 2,75 3,25 3 3,25 2,75 3,25 3 3,25 2,75 2,5 2,75 2,75 3 2,25 0

Step 2.2 Normalizing the direct relation matrix

The k value was calculated as 0.015 using Equation 3 and the average direct relationship matrix was normalized using Equation 4 with this value. The normalized direct relationship matrix is shown in Table 3.

Table 3

Normalized direct relation matrix [M]

C _i	C ₁	C ₂	C ₃	C ₄	C ₅	C ₆	C ₇	C ₈	C ₉	C ₁₀	C ₁₁	C ₁₂	C ₁₃	C ₁₄	C ₁₅	C ₁₆	C ₁₇	C ₁₈	C ₁₉	C ₂₀
C ₁	0,000	0,031	0,050	0,034	0,038	0,034	0,042	0,027	0,042	0,046	0,034	0,038	0,034	0,023	0,031	0,0307	0,0421	0,046	0,0268	0,0268
C ₂	0,050	0,000	0,046	0,046	0,023	0,023	0,046	0,054	0,046	0,027	0,027	0,034	0,050	0,027	0,027	0,0307	0,0383	0,0345	0,0268	0,023
C ₃	0,015	0,015	0,000	0,015	0,015	0,011	0,027	0,031	0,019	0,023	0,019	0,019	0,019	0,023	0,019	0,0192	0,0307	0,0345	0,023	0,023
C ₄	0,023	0,023	0,042	0,000	0,019	0,023	0,027	0,034	0,031	0,023	0,019	0,034	0,027	0,019	0,023	0,0153	0,0345	0,0383	0,0192	0,0192
C ₅	0,061	0,061	0,057	0,061	0,000	0,054	0,061	0,057	0,061	0,061	0,057	0,061	0,061	0,038	0,034	0,0383	0,046	0,046	0,0345	0,0345
C ₆	0,057	0,061	0,061	0,061	0,057	0,000	0,061	0,061	0,061	0,061	0,057	0,061	0,061	0,038	0,034	0,0345	0,0498	0,0498	0,0345	0,0345
C ₇	0,023	0,031	0,034	0,027	0,015	0,011	0,000	0,019	0,023	0,034	0,034	0,023	0,019	0,019	0,027	0,0192	0,0345	0,0307	0,0192	0,0192
C ₈	0,046	0,023	0,046	0,038	0,015	0,019	0,042	0,000	0,046	0,050	0,042	0,031	0,031	0,019	0,023	0,023	0,0345	0,0345	0,023	0,023
C ₉	0,019	0,019	0,019	0,015	0,011	0,019	0,038	0,031	0,000	0,027	0,023	0,023	0,023	0,023	0,027	0,0192	0,0383	0,0345	0,0192	0,023
C ₁₀	0,011	0,019	0,038	0,046	0,019	0,015	0,042	0,023	0,031	0,000	0,023	0,027	0,031	0,019	0,027	0,023	0,0345	0,0383	0,023	0,023
C ₁₁	0,019	0,027	0,023	0,034	0,015	0,015	0,042	0,027	0,027	0,027	0,000	0,027	0,027	0,023	0,027	0,0268	0,0383	0,0345	0,0192	0,0192
C ₁₂	0,015	0,015	0,038	0,015	0,008	0,011	0,031	0,015	0,031	0,019	0,015	0,000	0,015	0,015	0,019	0,023	0,0307	0,0307	0,0192	0,0192
C ₁₃	0,019	0,019	0,038	0,019	0,019	0,015	0,023	0,023	0,038	0,023	0,023	0,023	0,000	0,027	0,019	0,023	0,0268	0,0268	0,0153	0,0153
C ₁₄	0,046	0,046	0,057	0,046	0,034	0,034	0,054	0,054	0,057	0,054	0,050	0,057	0,054	0,000	0,050	0,0383	0,0498	0,0498	0,0268	0,0268
C ₁₅	0,034	0,038	0,038	0,038	0,019	0,019	0,038	0,031	0,023	0,023	0,023	0,038	0,038	0,019	0,000	0,0192	0,0383	0,0345	0,0192	0,0192
C ₁₆	0,046	0,042	0,046	0,042	0,031	0,027	0,046	0,038	0,046	0,046	0,046	0,046	0,046	0,027	0,046	0	0,046	0,046	0,0268	0,0268
C ₁₇	0,023	0,023	0,038	0,023	0,019	0,019	0,038	0,023	0,023	0,023	0,023	0,023	0,023	0,015	0,023	0,023	0	0,023	0,023	0,023
C ₁₈	0,034	0,023	0,042	0,023	0,023	0,023	0,042	0,027	0,027	0,027	0,027	0,023	0,027	0,019	0,023	0,0115	0,0383	0	0,0192	0,0192
C ₁₉	0,046	0,050	0,046	0,046	0,050	0,046	0,054	0,046	0,050	0,050	0,046	0,050	0,046	0,042	0,042	0,0383	0,0498	0,046	0	0,0383
C ₂₀	0,054	0,042	0,042	0,038	0,042	0,042	0,050	0,046	0,050	0,042	0,050	0,046	0,050	0,042	0,038	0,0421	0,0421	0,046	0,0345	0

Step 2.3 Creating the total relation matrix

Total relation matrix is obtained from the normalized direct relation matrix by using Equation 5 as shown in Table 4.

Table 4

Total relation matrix [T]

c _i	c ₁	c ₂	c ₃	c ₄	c ₅	c ₆	c ₇	c ₈	c ₉	c ₁₀	c ₁₁	c ₁₂	c ₁₃	c ₁₄	c ₁₅	c ₁₆	c ₁₇	c ₁₈	c ₁₉	c ₂₀
c ₁	0,051	0,079	0,114	0,087	0,075	0,071	0,106	0,08	0,099	0,0997	0,085	0,0919	0,0879	0,0619	0,0755	0,0705	0,102	0,1038	0,064	0,0643
c ₂	0,098	0,047	0,11	0,097	0,06	0,059	0,109	0,104	0,102	0,0811	0,077	0,0872	0,1011	0,0648	0,0712	0,0698	0,0975	0,0922	0,0633	0,06
c ₃	0,047	0,045	0,04	0,048	0,038	0,034	0,066	0,062	0,055	0,0563	0,0504	0,052	0,0518	0,0463	0,0468	0,0436	0,067	0,0694	0,0453	0,0455
c ₄	0,06	0,057	0,089	0,039	0,046	0,049	0,074	0,072	0,072	0,0626	0,056	0,0728	0,0652	0,0472	0,0554	0,0445	0,0777	0,08	0,0462	0,0464
c ₅	0,132	0,128	0,15	0,136	0,055	0,105	0,153	0,132	0,144	0,1384	0,129	0,1373	0,1368	0,0935	0,099	0,0957	0,1323	0,13	0,0877	0,088
c ₆	0,129	0,129	0,154	0,137	0,11	0,054	0,154	0,137	0,144	0,1392	0,1297	0,138	0,1375	0,0941	0,0996	0,0927	0,1367	0,1343	0,0882	0,0886
c ₇	0,057	0,062	0,078	0,063	0,041	0,037	0,045	0,055	0,062	0,0706	0,068	0,0594	0,0557	0,0454	0,0572	0,0465	0,0751	0,0702	0,0444	0,0446
c ₈	0,088	0,064	0,102	0,084	0,048	0,051	0,098	0,046	0,095	0,0961	0,0852	0,0771	0,0766	0,0531	0,0625	0,058	0,0871	0,0857	0,0554	0,0558
c ₉	0,053	0,051	0,063	0,051	0,037	0,044	0,081	0,065	0,039	0,0634	0,0573	0,0591	0,0589	0,0487	0,057	0,0461	0,0782	0,0732	0,0441	0,0479
c ₁₀	0,049	0,055	0,086	0,084	0,047	0,043	0,089	0,062	0,073	0,0408	0,0605	0,0666	0,0698	0,048	0,0601	0,0524	0,079	0,0812	0,0504	0,0507
c ₁₁	0,056	0,061	0,071	0,073	0,043	0,042	0,088	0,065	0,069	0,0663	0,0375	0,0659	0,0656	0,0509	0,0596	0,0556	0,0818	0,0767	0,0462	0,0464
c ₁₂	0,044	0,042	0,074	0,045	0,029	0,032	0,067	0,045	0,062	0,0499	0,0441	0,0306	0,0455	0,0373	0,0448	0,0453	0,0645	0,0634	0,0401	0,0403
c ₁₃	0,052	0,05	0,08	0,053	0,043	0,039	0,065	0,057	0,075	0,0584	0,0558	0,0578	0,0351	0,0515	0,0486	0,0488	0,0659	0,0647	0,0396	0,0398
c ₁₄	0,109	0,105	0,139	0,113	0,081	0,08	0,135	0,119	0,13	0,1216	0,1129	0,1243	0,1202	0,0497	0,1059	0,0886	0,1258	0,1235	0,074	0,0744
c ₁₅	0,074	0,075	0,09	0,08	0,049	0,048	0,089	0,073	0,07	0,0665	0,0633	0,0806	0,0803	0,0501	0,0364	0,0514	0,086	0,0809	0,049	0,0492
c ₁₆	0,101	0,094	0,117	0,1	0,072	0,067	0,117	0,096	0,109	0,1053	0,1009	0,1047	0,1043	0,0695	0,0949	0,0451	0,1122	0,1102	0,0679	0,0682
c ₁₇	0,057	0,055	0,082	0,059	0,045	0,044	0,081	0,059	0,062	0,0601	0,0575	0,0595	0,0592	0,0419	0,0535	0,0501	0,0415	0,0628	0,048	0,0482
c ₁₈	0,071	0,058	0,089	0,062	0,05	0,05	0,089	0,065	0,069	0,067	0,064	0,0625	0,0658	0,0477	0,0559	0,0415	0,082	0,0437	0,0466	0,0468
c ₁₉	0,112	0,112	0,131	0,116	0,098	0,094	0,138	0,115	0,126	0,1211	0,1123	0,1203	0,1163	0,0923	0,1008	0,0908	0,1283	0,1225	0,0497	0,087
c ₂₀	0,116	0,102	0,123	0,105	0,089	0,088	0,13	0,112	0,122	0,1106	0,1128	0,1133	0,1164	0,0899	0,0945	0,0919	0,1174	0,1189	0,0806	0,0477

Step 2.4 Determining the sender and receiver groups

The sender and receiver groups were determined by using Equations 6 –8 as shown in Table 5. According to this table, C₁, C₂, C₅, C₆,C₁₄, C₁₆, C₁₉ve C₂₀ criteria affect other criteria, i.e., they are in the sender group while the others are in the receiver group.

Table 5

Sender and receiver groups

C _i	D _i	R _i	D_i + R_i	D_i - R_i
C ₁	1,669	1,557	3,226	0,113
C ₂	1,652	1,475	3,126	0,177
C ₃	1,010	1,981	2,991	–0,972
C ₄	1,212	1,631	2,843	–0,419
C ₅	2,403	1,157	3,559	1,246
C ₆	2,427	1,132	3,559	1,295
C ₇	1,137	1,973	3,110	–0,836
C ₈	1,469	1,624	3,093	–0,155
C ₉	1,120	1,780	2,900	–0,660
C ₁₀	1,248	1,675	2,923	–0,427
C ₁₁	1,221	1,559	2,780	–0,339
C ₁₂	0,947	1,661	2,608	–0,714
C ₁₃	1,081	1,650	2,731	–0,569
C ₁₄	2,132	1,184	3,316	0,948
C ₁₅	1,343	1,379	2,722	–0,037
C ₁₆	1,857	1,229	3,086	0,628
C ₁₇	1,126	1,838	2,964	–0,712
C ₁₈	1,227	1,788	3,015	–0,560
C ₁₉	2,183	1,131	3,314	1,052
C ₂₀	2,080	1,140	3,220	0,940

Step 2.5. Determining the criteria weights

In this step, weight and relative weight values were calculated by using Equations 9–10 as shown in Table 6.

Table 6

Weight and relative weight values of the criteria

C _i	w _i	$W_{i}^{*}$	Rank
C ₁	3,228	0,051	6
C ₂	3,131	0,050	10
C ₃	3,145	0,050	9
C ₄	2,874	0,046	16
C ₅	3,771	0,060	2
C ₆	3,787	0,060	1
C ₇	3,220	0,051	7
C ₈	3,097	0,049	11
C ₉	2,975	0,047	14
C ₁₀	2,954	0,047	15
C ₁₁	2,800	0,045	17
C ₁₂	2,704	0,043	20
C ₁₃	2,790	0,044	18
C ₁₄	3,449	0,055	4
C ₁₅	2,722	0,043	19
C ₁₆	3,149	0,050	8
C ₁₇	3,049	0,049	13
C ₁₈	3,066	0,049	12
C ₁₉	3,477	0,055	3
C ₂₀	3,355	0,053	5

Table 6 shows that the most important criterion is lower back strain criteria, which is followed by upper back strain and time pressure criteria. Femur strain was found to be the least important criterion.

Stage 3. Ranking the alternatives by using TOPSIS Method

At this stage, the AMTs with license categories A, B1, B2, and B1 + B2 were compared on the basis of workloads and ranked according to workloads.

Step 3.1. Creating the decision matrix

In this step, four DMs were asked to evaluate license categories according to workloads. Since these evaluations were quite similar, the decision matrix displayed in Table 7 was created according to the common views of the DMs by using Equation 11. 1-9 rating scale was used.

Table 7

Decision matrix [R]

Criteria/Category	C ₁	C ₂	C ₃	C ₄	C ₅	C ₆	C ₇	C ₈	C ₉	C ₁₀	C ₁₁	C ₁₂	C ₁₃	C ₁₄	C ₁₅	C ₁₆	C ₁₇	C ₁₈	C ₁₉	C ₂₀
A	7	6	6	7	7	7	6	5	5	6	6	6	7	7	6	6	7	6	8	8
B1	8	7	7	7	9	9	7	8	6	7	7	6	8	9	8	9	8	7	9	9
B2	7	6	7	7	8	8	6	7	6	6	6	6	7	9	7	8	8	6	9	8
B1 + B2	8	8	8	8	9	9	7	8	6	7	7	7	8	9	8	9	8	7	9	9

Step 3.2. Normalizing the decision matrix matrix, which is displayed in Table 7, was normalized using Equation 12, and the normalized [N] matrix was created.

Step 3.3. Calculating the weighted normalized decision matrix

[V] is obtained by multiplying $W_{i}^{*}$ values, which were calculated using the DEMATEL method and displayed in Table 6, with the normalized values. Table 8 presents the [V] matrix.

Table 8

Weighted normalized decision matrix [V]

Criteria/Category	C ₁	C ₂	C ₃	C ₄	C ₅	C ₆	C ₇	C ₈	C ₉	C ₁₀	C ₁₁	C ₁₂	C ₁₃	C ₁₄	C ₁₅	C ₁₆	C ₁₇	C ₁₈	C ₁₉	C ₂₀
A	0,024	0,022	0,021	0,022	0,025	0,025	0,024	0,017	0,021	0,022	0,021	0,021	0,021	0,023	0,018	0,019	0,022	0,022	0,025	0,025
B1	0,027	0,026	0,025	0,022	0,033	0,033	0,028	0,028	0,025	0,025	0,024	0,021	0,024	0,029	0,024	0,028	0,025	0,026	0,028	0,028
B2	0,024	0,022	0,025	0,022	0,029	0,029	0,024	0,024	0,025	0,022	0,021	0,021	0,021	0,029	0,021	0,025	0,025	0,022	0,028	0,025
B1 + B2	0,027	0,029	0,028	0,025	0,033	0,033	0,028	0,028	0,025	0,025	0,024	0,024	0,024	0,029	0,024	0,028	0,025	0,026	0,028	0,028

Step 3.4. Calculating the positive ideal and negative ideal solutions

Since all the criteria in the present study are cost type criteria, PIS and NIS were calculated by using Equation 16 as shown in Table 9.

Table 9

PIS and NIS values

Criteria/Solution	C ₁	C ₂	C ₃	C ₄	C ₅	C ₆	C ₇	C ₈	C ₉	C ₁₀	C ₁₁	C ₁₂	C ₁₃	C ₁₄	C ₁₅	C ₁₆	C ₁₇	C ₁₈	C ₁₉	C ₂₀
PIS	0,024	0,022	0,021	0,022	0,025	0,025	0,024	0,017	0,021	0,022	0,021	0,021	0,021	0,023	0,018	0,019	0,022	0,022	0,025	0,025
NIS	0,027	0,029	0,028	0,025	0,033	0,033	0,028	0,028	0,025	0,025	0,024	0,024	0,024	0,029	0,024	0,028	0,025	0,026	0,028	0,028

Step 3.5. Calculating the distances of the alternatives to PIS and NIS

Equations 18 were used to calculate $S_{t}^{*}$ and $S_{t}^{-}$ values as shown in Table 10.

Step 3.6. Determining the relative closeness to ideal solution and ranking alternatives

In this step, C_t* value of each alternative was calculated by using Equation 19 as shown in Table 10, and alternative license categories were sorted according to these values.

Table 10

Distance values according to PIS and NIS and Relative closeness values of the alternatives

Category	$S_{t}^{*}$	$S_{t}^{-}$	$C_{t}^{*}$	Rank
A	0,000	0,021	1,000	1
B1	0,023	0,007	0,232	3
B2	0,015	0,014	0,485	2
B1 + B2	0,025	0,000	0,000	4

According to Table 10, the license category with the lowest risk is the A category and the license category with the highest risk is the B1 + B2 category in terms of total workload. Preventive and regulatory precautions should be implemented to reduce the workloads of AMTs holding the B1 + B2 license category.

5 Discussion and conclusion

The current study proposed a new hybrid MCDM approach for the evaluation of AMTs in terms of total workloads by taking into consideration of physical, mental, and environmental workload criteria since they are likely to suffer from physical workloads stemming from working posture as well as mental workloads due to intense and stressful working hours and environmental workloads due to environmental conditions. The proposed hybrid approach revealed that strain on the upper back and the lower back, the time pressure, and the air temperature should be prioritized over other workload criteria according to the DEMATEL method. The studies focusing on musculoskeletal system disorders due to the physical workloads of aircraft maintenance technicians [38 , 40–43] similarly reported that the upper back and lower back are the areas that are most strained. It is possible to explain the finding which reports that time pressure is one of the most significant criteria by the demands of managers in the aviation sector for shorter maintenance time. This finding should be taken into consideration so that time-related demands can be managed more effectively by the maintenance organization in relation to the following issues: expertise and limitations of technicians, work intervals, shift handovers, available personnel, setting deadlines, and prioritizing and allocating work tasks [78]. In addition, AMT staff is exposed to temperatures reaching 45°C in summer and –20°C in winter since they work outdoors in aprons. These temperatures are felt higher when combined with humidity, which makes working conditions considerably more challenging. Moreover, weather conditions such as rain, snow, and storm are among other factors leading to more challenging working conditions for technicians working outdoors.

Lifting heavy objects that strain AMTs’ musculoskeletal system and push their abilities to the limits might result in injuries. It might be easier to fulfill tasks that are beyond the physical limits of the human body when supplementary machines or tools such as a crane are used to lift an aircraft. The design of modern aircraft mostly facilitates maintenance procedures by allowing technicians to access aircraft equipment easily and apply an adequate amount of force to tighten or loosen screws. However, such solutions might not be sufficient to prevent AMTs from getting injured [8, 22], which might explain the finding of the present study which claims that the most important criteria for line maintenance technicians are strain in the upper back and lower back. Unlike technicians working in other maintenance areas, those working in line maintenance carry out tasks requiring lifting since they are responsible for mounting certain spare parts to aircraft in the apron. Such frequent lifting experiences are known to cause lower backaches [5, 44]. In addition, line maintenance technicians stand for a long time on their feet, which also contributes to upper backache and lower backache. The related studies concluded that lower back ache symptoms might be observed when one stands on his feet for more than 71 minutes [78, 79]. Long working hours might also result in aches in the arm, legs, neck, and back.

According to the results of the TOPSIS method, the license categories of AMTs were sorted according to the workloads as follows: A, B2, B1, and B1 + B2. It is a great advantage for aircraft maintenance organizations to employ AMTs holding B1 + B2 licenses since these two category licenses allow AMTs to work in line maintenance, be promoted to flying technician positions, or be assigned duties abroad when necessary. An AMT holding these two license categories can prepare an aircraft for flight more easily, which is a great advantage for the organization both in terms of time and cost. However, this advantage is likely to lead to extreme mental and physical workloads, risk technicians’ health, and demotivate them. In addition, environmental factors might also challenge the exhaustion limits of technicians and negatively affect their fault-detection times. Therefore, both strictly following the schedules and plans specified in task cards and the presence of a sufficient number of AMTs working on maintenance tasks are critical to ensure flight safety. In addition, workloads might be minimized by providing the staff with suitable and necessary equipment for the planned tasks. In other words, aircraft maintenance organizations should take the findings of the present study into consideration to ensure flight safety by revising the salary and promotion procedures of the AMTs who deal with more challenging tasks and by taking necessary precautions regarding equipment and staff to minimize such challenges. Allocation of tasks and delegation of duties also play significant roles in boosting the work performances of AMTs [5].

When employees work for a long time on a task without changing their body posture, they suffer from recurring aches because the body weakens in terms of body forms and elasticity in time. Therefore, it is essential to make necessary revisions and rearrangements in the tasks and working environments by employing Ergonomic analysis methods to minimize workloads and prevent health problems due to long working hours, prolonged stress, messy working environments, and improper environmental factors [3]. Therefore, AMTs might be informed about exercise and ergonomic principles during on-the-job training practices. In addition, more studies might be conducted to determine the maximum amount of workload that technicians can tolerate. Moreover, the staff might be provided information regarding the use of proper tools and equipment for a specific work task. Finally, work uniforms that keep the body cool and are not affected by humidity might be preferred to alleviate the unfavorable effect of heat.

Selecting only 2 dimensions of the NASA TLX in this study can be considered a limitation of this study. If some of the other dimensions in the NASA-TLX were selected (such as mental demand or frustration level); the results of this study may be different. The reasons for taking only 2 dimensions in the study; The work done by the technicians in the line maintenance is mostly physical work. In addition, it was considered appropriate not to use the other dimensions of the NASA TLX in order to the results in the developed model, since the DMs in our study had difficulty responding to the comparisons involving other dimensions and were not fully understood. Aircraft maintenance engineering is a relatively active occupation. Regardless of the job being done, most tasks tend to have elements of fine motor control, requiring precision, as well as activities requiring strength and gross manipulation [80]. The participation of DMs working on different airlines may be effective in the results. Therefore, it would be appropriate to refer to the opinions of experts from different airlines in future studies.

Further studies might examine total workloads in workshops where AMTs work such as base maintenance, airframe, avionics, dyeing, coating and seats, and other related units by using different MCDM methods. Later, the most challenging tasks for AMTs as well as possible precautions to take might be determined according to these workloads.

Ethical approval

Not applicable.

Informed consent

Not applicable.

Conflict of interest

The authors have no conflict of interest to report.

Footnotes

Acknowledgments

This study presents the results of the Master of Science (MS) thesis of the first author, entitled “Hybrid Multi-Criteria Decision Making Approach Proposal for Evaluating Workloads of Aircraft Maintenance Technicians”.

Funding

None to report.

References

Becker

, Warm

, Dember

, Hancock

. Effects of Jet Engine Noise and Performance Feedback on Perceived Workload in a Monitoring Task. The International Journal of Aviation Psychology. 1995;5(1):49–62.

Dagdeviren

, Eraslan

, Kurt

. A Model and Its Application for Determining the Total Workload Levels of Employees. Gazi University Journal of the Faculty of Engineering and Architecture [Internet]. 2013;20(4):0.

Adar

, Delice

. Comparison of the total workload of local public transportation drivers with a new approach according to physical and mental workload criteria. Pamukkale Üniversitesi Mühendislik Bilimleri Dergisi [Internet]. 2020;26(1):254–67.

DHMI. 2020 Airline Industry Report. General Directorate Of State Airports Authority Website. 2021 [cited 2022 Feb 25]. Available from: https://www.dhmi.gov.tr/Lists/HavaYoluSektorRaporlari/Attachments/14/2020_Havayolu_Sektor_raporu.pdf

Civil Aviation Authority. CAP 715: An Introduction to Aircraft Maintenance Engineering Human Factors for JAR 66 [Internet]. 2002. Available from: https://publicapps.caa.co.uk/docs/33/CAP715.PDF

Skybrar. Continuing Airworthiness. 2017. Available from: https://www.skybrary.aero/index.php/Continuing_Airworthiness

Yazgan

, Kucuk Yilmaz

. Prioritisation of Factors Contributing to Human Error for Airworthiness Management Strategy with ANP. Aircraft Engineering and Aerospace Technology. 2019;91(1):78–93.

Asadi

, Yu

, Mott

. Risk Factors For Musculoskeletal Injuries In Airline Maintenance, Repair & Overhaul. International Journal Of Industrial Ergonomics. 2019;70:107–15.

Dhillon

, Liu

. Human Error in Maintenance: A Review. Journal of Quality in Maintenance Engineering. 2006;12(1):21–36.

10.

Babalık

. Ergonomics for Engineers (Business Science). Dora Publications. 2016.

11.

Yazgan

. Development Taxonomy of Human Risk Factors for Corporate Sustainability in Aviation Sector. Aircraft Engineering and Aerospace Technology. 2018;90(6):1012–22.

12.

Fontela

, Gabus

. DEMATEL, Innovative Methods. Report no. 2 Structural Analysis of the World Problematique, Battelle Geneva Research Institute. 1974;67-69.

13.

Yuan

, Xu

, Zhang

. The DEMATEL–COPRAS hybrid method under probabilistic linguistic environment and its application in Third Party Logistics provider selection. Fuzzy Optimization and Decision Making. 2021;21:137–56.

14.

Ghosh

, Das Chatterjee

, Dinda

. Urban ecological security assessment and forecasting using integrated DEMATEL-ANP and CA-Markov models: A case study on Kolkata Metropolitan Area, India. Sustainable Cities and Society. 2021;68:102773.

15.

Vishnu

, Sridharan

, Ram Kumar

, Regi Kumar

. Analysis of the operational risk factors in public hospitals in an Indian state. International Journal of Health Care Quality Assurance. 2019;33(1):67–88.

16.

Büyüközkan

, Güleryüz

. An integrated DEMATEL-ANP approach for renewable energy resources selection in Turkey. International Journal of Production Economics. 2016;182(C):435–48.

17.

Gölcük

, Baykasoğlu

. An analysis of DEMATEL approaches for criteria interaction handling within ANP. Expert Systems with Applications. 2016;46:346–66.

18.

Bongo

, Alimpangog

KMS

, Loar

, Montefalcon

, Ocampo

. An application of DEMATEL-ANP and PROMETHEE II approach for air traffic controllers workload stress problem: A case of Mactan Civil Aviation Authority of the Philippines. Journal of Air Transport Management [Internet]. 2018;68(C):198–213.

19.

Liou

JJH

, Yen

, Tzeng

G-H

. Building An Effective Safety Management System For Airlines. Journal Of Air Transport Management. 2008;14(1):20–6.

20.

Petrović

, Kankaras

. DEMATEL-AHP multi-criteria decision making model for the determination and evaluation of criteria for selecting an air traffic protection aircraft. Decision Making: Applications in Management and Engineering. 2018;1(2).

21.

Guo

, Chen

, Si

, Wang

. Evolution Mechanism on the Unsafe Behavioural Risks of General Aviation Pilots. Engineering Economics. 2021;32(2):104–17.

22.

Öztürk

, Yazgan

, Kılıc Delice

. Evaluation of Physical Workload Factors with DEMATEL Method for Aircraft Maintenance Technicians. European Journal of Science and Technology. 2021;(28):1447–53.

23.

Hwang

, Yoon

. Methods for Multiple Attribute Decision Making. Multiple Attribute Decision Making. 1981;58-191.

24.

Delice

, Can

, Kahya

. Improving the rapid office strain assessment method with an integrated multi-criteria decision making approach. Gazi University Journal of the Faculty of Engineering and Architecture. 2020;35(3):1297–314.

25.

Konstantinos

, Georgios

, Garyfalos

. A Decision Support System methodology for selecting wind farm installation locations using AHP and TOPSIS: Case study in Eastern Macedonia and Thrace region, Greece. Energy Policy. 2019;132:232–46.

26.

Azimifard

, Moosavirad

, Ariafar

. Selecting sustainable supplier countries for Iran’s steel industry at three levels by using AHP and TOPSIS methods. Resources Policy. 2018;57(C):30–44.

27.

Pourjavad

, Shirouyehzad

, Shahin

. Selecting maintenance strategy in mining industry by analytic network process and TOPSIS. International Journal of Industrial and Systems Engineering. 2013;15(2):171.

28.

Prasad

, Rajesh

, Thırumalaıkumarasamy

. Selection of coating material for magnesium alloy using Fuzzy AHP-TOPSIS. Sádhaná. 2020;45(1).

29.

Sari

. Forest Fire Susceptibility Mapping Via Multi-Criteria Decision Analysis Techniques for Mugla, Turkey: A comparative analysis of VIKOR and TOPSIS. Forest Ecology and Management. 2021;480:118644.

30.

Saitoglu

, Apak

. Multi-Dimensional Competency Assessment: An Application In Aircraft Maintenance Organization. Pressacademia. 2021;8(2):101–28.

31.

Ahmadı

, Gupta

, Karım

, Kumar

. Selection Of Maintenance Strategy For Aircraft Systems Using Multi-Criteria Decision Making Methodologies. International Journal of Reliability, Quality and Safety Engineering. 2010;17(03):223–43.

32.

Ahmad

, Ouenniche

, Kolosz

, Greening

, Andresen

, Maroto-Valer

, et al. A stakeholders’ participatory approach to multi-criteria assessment of sustainable aviation fuels production pathways. International Journal of Production Economics. 2021;238:108156.

33.

Wang

T-C

, Chang

T-H

. Application of TOPSIS in evaluating initial training aircraft under a fuzzy environment. Expert Systems with Applications. 2007;33(4):870–80.

34.

Erdinc

, Hot

, Ozkaya

. Turkish version of the Cornell Musculoskeletal Discomfort Questionnaire: Cross-cultural adaptation and validation. Work. 2011;39(3):251–60.

35.

DGCA. SHT-145 Approved Aircraft Maintenance Organizations Instruction [Internet]. 2018 [cited 2022 Feb 25]. Available from: https://web.shgm.gov.tr/documents/sivilhavacilik/files/mevzuat/sektorel/talimatlar/2018/SHT-145.pdf

36.

Tatar

, Yazicioglu

, Ayvaz

. A Novel Risk Assessment Model for Work-related Musculoskeletal Disorders in Tea HarvestingWorkers. Journal of Intelligent & Fuzzy Systems Work. 2022;1-19.

37.

Akay

, Kurt

, Dagdeviren

. Ergonomic Analysis of Working Positions. Gazi University Journal of the Faculty of Engineering and Architecture. 2013;18(3):73–84.

38.

Chae

, Kim

. Risk factors for musculoskeletal symptoms in aviation maintenance technicians. Korean Journal of Occupational and Environmental Medicine. 2005;17(3):173–85.

39.

Kemp

, Burnham

, Copley

, Shim

. Injuries to Air Force Personnel Associated with Lifting, Handling, and Carrying Objects. American Journal of Preventive Medicine. 2010;38(1):148–55.

40.

Nogueira

, Diniz

ACP

, Barbieri

, Padula

, Carregaro

, de Oliveira

. Musculoskeletal disorders and psychosocial risk factors among workers of the aircraft maintenance industry. Work. 2012;41:4801–7.

41.

Zungu

. A Comparative Study of the Prevalence and Risk Factors of Lower Back Pain among Aircraft Technicians in Ethiopian Airlines. Occupational Health Southern Africa. 2015;21(2):18–23.

42.

Irwin

, Streilein

. Use of Field-based Motion Capture to Augment Observational data in ergonomic Assessment of Aircraft Maintenance. Procedia Manufacturing. 2015;3:4501–8.

43.

Fajardo Rodriguez

, Ortiz Mayorga

. Characterization of Low Back Pain in Pilots and Maintenance Technicians on a Commercial Airline. Aerospace Medicine and Human Performance. 2016;87(9):795–9.

44.

Yazgan

, Ozkan

, Haktanırlar Ulutas

. A Questionnaire-Based Musculoskeletal Disorder Assessment for Aircraft Maintenance Technicians. Aircraft Engineering and Aerospace Technology. 2021;94(2):240–7.

45.

McCormick

, Sanders

. Human Factors in Engineering and Design (6th ed.), New York: McGraw-Hill. 1982.

46.

Wickens

, Gordon

, Liu

. An Introduction to Human Factors Engineering. New York: Longman. 1998.

47.

Mishra

, Amitabha

, Iqbal

, Gangopadhyay

, Chandra

. An integrative approach for evaluating work related musculoskeletal disorders. Work. 2012;43(4):437–46.

48.

Çilingir

. Ergonomic Studies on Human Workload and Efficiency in Some Work Done with Agricultural Machinery. Published Ph.D. Thesis, METU Department of Industrial Engineering. 1981.

49.

Can

, Delice

, Özcakmak

. Selection of Seating Arrangement by Using Multi Criteria Decision Making Approach. Journal of Engineering Sciences and Design. 2017;5:215–23.

50.

Crawford

. The Nordic Musculoskeletal Questionnaire. Occupational Medicine. 2007;57(4):300–1.

51.

Günes

, Ceylan

. Investigatıon of Ergonomic Conditions In Computerized Working Environment and Musculoskeletal Discomforts of Academician Nurses. Journal of Ege University Faculty of Nursing. 2016;32(2):61–74.

52.

Kahya

, Özkan

. Evaluation of ergonomic risks in administrative offices of a university. Journal of Gazi University Faculty of Engineering and Architecture. 2017;32(1):141–50.

53.

Gönen

, Karaoglan

, Ocaktan

MAB

, Oral

, Atici

, Kaya

. A new risk assessment approach in the analysis of musculoskeletal disorders. Journal of Gazi University Faculty of Engineering and Architecture [Internet]. 2018;33(2):425–40.

54.

Sugiharto

. The Relationship between Mental Workload and Occupational Stress among Aircraft Maintenance Officers at PT X. The Indonesian Journal of Occupational Safety and Health. 2019;8(2):233–9.

55.

Eggemeier

, Wilson

. Subjective and Performance-based Assessment ofWorkload in Multitask Environments. Multiple Task Performance/London. 217-278. 1991.

56.

Longo

. Formalizing Human MentalWorkload as a Defeasible Computational Concept.Doctor of Philosophy Thesis (Computer Science), University of Dublin, Trinity College. 2014.

57.

Yeh

, Wickens

. Why do performance and subjective measures dissociate? Human Factors Society 28th Annual Meeting/Santa Monica. 504-8. 1984.

58.

Hart

, Staveland

. Development of NASA-TLX (Task Load Index): Results of Empirical and Theoretical Research [Internet]. Hancock PA, Meshkati N, editors. Vol. 52, ScienceDirect. North-Holland; p. 139-83. 1988.

59.

Eraslan

, Can

, Atalay

. Mental Workload Assessment Using a Fuzzy Multi-criteria Method. Tehnivki vjesnik. 2016;23(3):667–74.

60.

ISO. ISO 10075-1. 2017. [cited 2022 May 28] Available from: https://www.iso.org/obp/ui/#iso:std:iso:10075:-1:ed-1:v1:en

61.

Hancock

, Caird

. Experimental Evaluation of a Model of Mental Workload. Human Factors: The Journal of the Human Factors and Ergonomics Society. 1993;35(3):413–29.

62.

Hill

, Byers

, Zaklad

, Christ

. Subjective Workload Assessment during 48 Continuous Hours of LOS-F-H Operations. Proceedings of the Human Factors Society Annual Meeting. 1989;33(16):1129–33.

63.

Hoonakker

, Carayon

, Gurses

, Brown

, Khunlertkit

, McGuire

, et al. Measuring workload of ICU nurses with a questionnaire survey: the NASA Task Load Index (TLX). IIE Transactions on Healthcare Systems Engineering. 2011;1(2):131–43.

64.

Oge

. Ergonomics ForOrganizational Effectivenes, Çizgi Bookstore, Konya, p. 189. 2015.

65.

W-W

. Choosing knowledge management strategies by using a combined ANP and DEMATEL approach. Expert Systems with Applications. 2008;35(3):828–35.

66.

Tzeng

, Chıang

, Lı

. Evaluating intertwined effects in e-learning programs: A novel hybrid MCDM model based on factor analysis and DEMATEL. Expert Systems with Applications. 2007;32(4):1028–44.

67.

Yılmaz Kaya

. Minimizing OHS Risks with Spherical Fuzzy Sets as a Verdict to Inventory Management: A Case Regarding Energy Companies. Discrete Dynamics in Nature and Society, 2022. 9511339. https://doi.org/10.1155/2022/9511339

68.

Adar

, Delice

. New integrated approaches based on MC-HFLTS for healthcare waste treatment technology selection. Journal of Enterprise Information Management. 2019;32(4):688–711.

69.

Yazgan

, Kucuk Yılmaz

. Prioritization of Factors Contributing To Human Error for Airworthiness Management Strategy with ANP. Aircraft Engineering and Aerospace Technology. 2019;91(1):78–93.

70.

Can

, Kılıc Delice

. A task-based fuzzy integrated MCDM approach for shopping mall selection considering universal design criteria. Soft Computing. 2018;22(9):7377–97.

71.

Delice

, Can

. A new approach for ergonomic risk assessment integrating KEMIRA, Best–Worst and MCDM methods. Soft Computing. 2020;24(19):15093–110.

72.

Karadağ

, Delice

. A New Fuzzy Hybrıd MCDM-ZOGP Approach For Product Selectıon Problem: A Case Study Of The Optıcal Fıber Cable. Internatıonal Journal of Industrıal Engıneerıng-Theory Applıcatıons and Practıce. 2018;25(4):507–25.

73.

Delice

, Can

. A stochastic approach for failure mode and effect analysis. RAIRO-Operations Research. 2017;51(4):1077–100.

74.

Haghighat

. Evaluating airline service quality using fuzzy DEMATEL and ANP. Strategic Public Management Journal. 2017;3(6):57–77.

75.

Saaty

. Rank from comparisons and from ratings in the analytic hierarchy/network processes. European Journal of Operational Research. 1987;30(3):228–36.

76.

Ishizaka

, Labib

. Review of the main developments in the analytic hierarchy process. Expert Systems with Applications. 2011;38(11):14336–45.

77.

DGCA. Aircraft Maintenance Personnel License Regulation (SHY-66) [Internet]. 2013 [cited 2022 Feb 25]. Available from: https://www.mevzuat.gov.tr/File/GeneratePdf?mevzuatNo=18972&mevzuatTur=KurumVeKurulusYonetmeligi&mevzuatTertip=5

78.

Kelsey

, Golden

, Mundt

. Low back pain/prolapsed lumbar intervertebral disc. Rheumatic Diseases Clinics of North America [Internet]. 1990;16(3):699–716.

79.

Riihimäki

. Its Origin and Risk Indicators. Scandinavian Journal of Work, Environment & Health. Scandinavian Journal of Work. Environment & HealthFinnish Institute of Occupational Health Danish National Research Centre for the Working Environment Norwegian National Institute of Occupational Health. 1991.

80.

CAA, Civil Aviation Authority. CAP 715: An introduction to aircraft maintenance engineering Human Factors for JAR 66. 2022. Available from: https://publicapps.caa.co.uk/docs/33/CAP715.PDF