Requirements for the Preliminary Design of Innovative Temporary Edge Protection Systems (TEPS) for Construction Works

Abstract

At present, it is striking that a large percentage of occupational accidents in the construction sector are still caused by falls from height. Therefore, curbing the severe personal, social and economic consequences of these events is not only a commitment but an obligation for all stakeholders in the construction sector. After a review of current fall protection systems on construction sites, the purpose of this study is to establish the preliminary requirements for the design, development and prototyping of a new system which can be used as an auxiliary means to prevent occupational accidents in the construction sector caused by fall hazards at height. Based on the design science research (DSR) methodology, this paper tests the capability of alternative materials (metals, plastics and composites) to withstand the loads required by the regulatory standard UNE-EN 13374:2013+A1:2019 and looks at the improvements they can offer. The results obtained enable new metals and composite materials to be put forward, based on their suitability to the parameters of the risks of falling from height, ensuring that the greatest number of potential situations are addressed. Then, the needs to be satisfied and requirements to be met are listed, prioritised and considered for new temporary edge protection systems (TEPS). Next, the attributes that increase user satisfaction and/or reduce user dissatisfaction are filtered by means of a Kano model, which is applied thanks to the responses of construction designers, coordinators and supervisors. Once these questions are solved, an analytic hierarchy process (AHP) is performed by a focus group, weighing the Kano contributions and ranking the materials to be selected for the preliminary design of innovative TEPS for construction works. After considering safety, ergonomics, adaptability, sustainability, efficiency, manufacturability and flexibility criteria, the basis for the design of a new temporary edge protection system is established.

1. Introduction

Construction has been essential for human survival throughout history. The archaeological, architectural and civil remains that exist all over the world give rise to the unsettling question of the working conditions that the labourers endured. We must ask ourselves, from what moment did humanity opt for the defence of life in the workplace [1]? Some renowned authors have ventured to formulate their own theories, fixing that moment to the publication of a text by Strabo [2]. However, even when taking into account the evolution of technical means and the specialisation of today’s workforce, safety conditions may not differ very much from those that existed in the past [3]. Any theory about the evolution of work and its conditions points to the fact that the concept of “safety” must have had different meanings throughout history. However, despite these changing conditions and the context of each era, construction has not slowed down.

Among the most popular topics in construction safety, accidents resulting from falls from height have garnered the attention of researchers over recent years [4], with numerous studies reflecting concern in this field [5,6,7]. Falls from height are the most frequent cause of injuries and deaths. The most common sources are: scaffolding or platforms without railings, workers not wearing safety harnesses correctly, unsound roofs, deficiencies in maintenance, the placement and securing of ladders, etc. These accidents have enormous consequences for people, business owners and society. Priority should, therefore, be given to measures that eliminate or reduce the risk at source and provide collective protection [8] over individual protection [9].

Domino theory [10], pioneering in the study of the causes of accidents (context, human error, unsafe conditions, accident, injury), indicates that 88% of accidents are caused by unsafe acts or behaviours, 10% are due to unsafe working conditions and the remaining 2% stem from unforeseeable causes. In this regard, several authors have pointed out that 90% of work accidents occur as a result of unsafe acts on the part of workers [11,12,13], confirming hypotheses put forward 60 years ago. However, other authors added that unsafe behaviours carried out by workers persist because they are often naturally reinforced [14]. For example, it may be difficult to properly use collective protection equipment with the auxiliary means available for carrying out tasks, as in the case of the outer railings on scaffolding trestles (sawhorses) [15].

Many unsafe behaviours of workers are reinforced positively (increased productivity) or negatively (reduced time and effort), which contributes to the increased likelihood that they will perform these same behaviours again in similar circumstances. In this sense, it is clear why operators engage in unsafe work behaviours [16], and the issue is simultaneously tackled by means of accident causality theories and human error theories [17]. Although workers are aware of the dangers associated with not wearing a safety harness or not having protective equipment, many forget or purposely do not use them when working at heights [18], so falls from height remain the main contributors to injuries and deaths on construction sites.

When analysing the time involved in assembling a security system for very small or one-off jobs, the arguments are the same, since there are no protection elements on the market that adapt to the specific circumstances of these small jobs, which, in many cases, only last a few hours or even minutes. Likewise, there are frequent complaints from operators that protective equipment, specifically, personal protective equipment (PPE), tends to hinder their movements, restricting their freedom and preventing or complicating the performance of specific tasks that take up very little time. This context shows that, despite having collective and/or individual protections, these complaints are not handled adequately. Therefore, to avoid this type of accident, and regardless of the behaviour related to the human factor, it is necessary to look for new methods that contribute to improving safety and health and reducing or eliminating subjectivity in production processes [19].

One of the most important measures to prevent falls from height on construction sites is the use of temporary edge protection systems (TEPS) [20,21]. Although studies of TEPS have been carried out from the point of view of the legal regulatory framework, based on compliance with a series of structural requirements [22], on the identification and evaluation of new needs, on innovation [23] or on the incorporation of new solutions and/or materials [24], no studies have been found that bring together all these considerations to address the innovative design, development and prototyping of new TEPS.

The rest of this paper is structured as follows. Section 2 presents the purpose of the research. Section 3 describes the methods selected to be followed based on the design science research (DSR) methodology (analysis of the current scenario by literature review, study of the behaviour of current collective protections by finite-element calculation, ranking of the attributes to be considered using the Kano model and proposal for the requirements of a preliminary design by means of an analytic hierarchy process (AHP)). Section 4 presents the results obtained. Section 5 discusses the findings. Finally, in Section 6, the main conclusions are presented.

2. Objectives

The objective of this study is to provide the basis of the preliminary design for the subsequent development and prototyping of new collective protection systems, focusing on adapting them to the parameters of the risk of falling from height and securing the greatest number of situations possible. As summarised in Figure 1, the new TEPS must be manufactured at a reasonable cost, ergonomic so that they are easy to assemble/disassemble, adaptable so that they can be used as auxiliary means in conjunction with other systems, flexible so that they can be used in unique situations, sustainable so that they have an extended service life, efficient to lighten the systems and safe to avoid occupational accidents due to the risk of falls from height, among other things. For this reason, the choice of materials used for their manufacture is crucial in order to incorporate all these characteristics. Therefore, this study aims to analyse the technical characteristics of the current systems manufactured with the current materials compared to a series of alternative materials and to subsequently propose the preliminary design of a new collective protection system capable of satisfying the set of needs detected and requirements listed, including safety as a key factor for their use. These objectives are aligned with Sustainable Development Goals 8 (Decent Work and Economic Growth) and 9 (Industry, Innovation and Infrastructure).

3. Methodology

The methodology followed is the DSR methodology, the objective of which is to develop innovative solutions to solve real problems while aiming to advance theoretical knowledge [25] by generating knowledge through the development of an innovative tool that solves a practical problem [26]. DSR starts with a descriptive phase for a deep understanding of the problem, and this gives way to a prescriptive phase [27]. DSR is a systematic yet flexible methodology aimed at improving engineering practices through iterative analysis, design, development and implementation based on collaboration between researchers and practitioners in real-world environments and leading to context-sensitive design theories and principles [28].

The practical problem addressed by this study is the selection of a series of alternative materials that can be used to lay the foundations for an innovative design of TEPS. It also considers the needs to be satisfied and the requirements to be met, listed, prioritised and hierarchised. This methodology has been successfully tested in the field [29,30]. This study, based on a review of the literature published on the subject and information obtained in the field (thanks to the authors’ professional experience in this field), analyses the needs and requirements that TEPS must meet through the Kano model [31], successfully tested in this field [32,33], in order to understand the reasons why workers in the construction sector fall from heights. Different materials are studied for the preliminary design of TEPS in relation to safety, efficiency and flexibility during their design and use as a minimum. The AHP tool [34,35] is used to hierarchise these alternative materials and recommend possible changes to accommodate aspects of sustainability, ergonomics, adaptability and manufacturability, among others. This combination of the Kano model with the AHP tool has been successfully tested both in product design in general [36,37] and specifically in terms of safety [38,39].

The Kano model is used to measure the effectiveness of products based on the identification of their requirements. The method classifies design preferences into five categories using a questionnaire that includes functional and dysfunctional questions. On the one hand, it considers the level of performance of the product and, on the other hand, the level of satisfaction of customers and/or users, which makes it possible to evaluate requirements by classifying them into 5 groups: mandatory, attractive, one dimensional, indifferent and reverse [40]. In this context, a requirement is considered mandatory if its absence causes dissatisfaction, even if its presence is taken for granted, and, therefore, its inclusion is not especially valued, while it is called attractive if customers and/or users value it when it is present, even if they do not notice its absence. Likewise, a requirement is called one dimensional if it increases satisfaction in proportion to the increase in its functionality and indifferent if customers and/or users are not interested or, secondly, if it has neutral and/or low impact. Finally, if the requirement undermines the functionality of the product and/or causes dissatisfaction, it is called reverse [41].

The AHP tool responds to the general approach of multi-criteria analysis for decision making. It is a discrete, multi-criteria decision-making method in which the problem to be solved is modelled from a set of alternatives and a series of decision criteria that, at times, can be conflicting [42]. The AHP is applied by building hierarchical structures in which the first hierarchical level consists of establishing an objective, the next hierarchical level, the decision criteria and, finally, the alternatives. For hierarchical levels, pairwise comparisons are made, either by assigning an absolute scale numerical value from 1 to 9 or by assigning a natural measurement scale value [43]. These comparisons result in dominance matrices. To apply AHP, it is necessary to analyse the alternatives that respond to the previously defined objectives. Next, the decision criteria must be selected. Then, those criteria are considered. Subsequently, the alternatives must be evaluated (after the level of satisfaction of each criterion). Once these steps have been taken, we proceed to the analytical resolution of the problem. Finally, the sensitivity of the decision is checked to measure the robustness of the proposal. This communication, which covers the first two phases, is part of the following framework:

• Phase 1. The first phase consists of the following stages:

  1.

  Analysis of the problem of occupational accidents in the construction sector, underlining the need to prevent falls from height;

  2.

  Description of current collective protection equipment, indicating dimensions, materials used, versatility and approximate assembly/disassembly time;

• Phase 2. The second phase consists of the following stages:

  3.

  Study of the behaviour of the collective protection equipment available on the market made with different materials;

  4.

  Analysis of the attributes that must be added into the new collective protection equipment to be designed, valuing basic, desired and motivating qualities;

• Phase 3. The third phase consists of the following stages:

  5.

  Hierarchisation of the attributes that must be incorporated into the new collective protection equipment to be designed;

  6.

  Proposal of alternative materials according to the requirements considered;

• Phase 4. The fourth phase consists of the following stages:

  7.

  Preliminary design of TEPS, taking into account their geometric and dimensional definition in order to study their patentability or usefulness;

  8.

  Detailed design and development of TEPS, considering all the previous steps;

• Phase 5. The fifth phase consists of the following stages:

  9.

  Prototyping TEPS, carrying out the corresponding laboratory tests to certify their technical suitability;

  10.

  Validation of solutions for their potential commercialisation, conducting market research surveys of users.

3.1. Analysis of the Problem of Occupational Accidents in the Construction Sector

The construction sector has a direct impact on a country’s economy and, therefore, plays an important role in its growth. However, on the other hand, it is a dangerous activity due to the high accident and death rates, as shown by alarming statistics [44]. On an international level, numerous research studies have reflected the high occupational accident rate caused by falls from height in the construction sector, exposing alarming rates of accidents and deaths, their costs and their causes [45,46,47,48]. In addition, fatal accidents frequently occur in building construction activities due to their inherently dangerous nature [49]. However, most accidents caused by falls from height happen because the risk of falling from relatively low heights is often underestimated. Therefore, a common factor in these accidents comes from the construction operators’ perception of the risk of falling since most fall accidents occur at elevations of less than 30 feet [20].

The Occupational Safety and Health Administration (OSHA) analysed 3496 deaths between 1985 and 1989 [50], showing that 33% of deaths were due to accidents caused by falls from height. Furthermore, OSHA concluded that deficiencies in protection equipment against falls from height also represent the highest number of claims, with injuries costing more than USD 5 trillion annually. These findings were confirmed by Cattledge et al. [51]. After that, Halabi et al. [52] listed 23,057 accidents recorded in the OSHA database in the years 2000–2020, stating that the proportion of accidents due to falls increased considerably and that the use of fall protection equipment had not improved. In Canada, the Canadian Centre for Occupational Health and Safety (CCOHS) also published that, of the 24,999 injuries caused in the construction sector between 1995 and 1998, 4676 were due to falling from height [53]. In addition, Winge and Albrechsen studied 176 accidents in the construction sector [54], which were investigated by the Norwegian Labour Inspection Authority in 2015, showing that many accidents could be explained by the lack of physical barrier elements, such as TEPS.

On the other hand, Zlatar et al. [55] analysed 114 cases of accidents due to falls from height from other research works, showing that, in 98% of the cases analysed, these were the result of inadequate or absent protection systems in the work procedures. Finally, the company XSPlatforms analysed the official information published in different countries on accidents caused by falls from height [56]:

• In the United States and Canada, falls from height amount to 36.9% of total accidents in the construction sector, being the number 1 cause in this sector. In addition, 3 out of 5 cases occur from a distance equal to or less than 20 feet;
• In France, accidents due to falls from height are the second leading cause of death at work, occupying the first position of occupational accidents in the construction sector (16%). In addition, within this sector they account for 30% of fatal accidents;
• In the UK, accidents due to falls from height within the construction sector account for 45% of fatal accidents. In addition, these are the most common cause of death and account for 3 out of 10 serious injuries;
• In Spain, falls from height are the most common cause of death in work accidents. Statistically, 1 in 20 accidents occurs as a result of a fall from height, half of these due to a fall from less than 3 m.

Table 1. Employment rates, workplace accidents and falls in Spain, 2000–2019. Source: INE.

Year	All Economic Sectors (General)			Construction Sector
	Occupancy	Accidents	Falls	Permits	Occupancy	Accidents	Falls
2000	16,146,275	932,932	163,758	121,246	1,695,900	239,244	44,591
2001	16,790,100	946,600	167,043	112,883	1,952,726	250,277	46,554
2002	17,475,600	938,188	164,806	116,903	2,189,274	250,414	46,618
2003	18,142,250	874,724	121,029	130,422	2,310,523	230,735	44,658
2004	19,207,000	871,724	128,829	146,408	2,455,722	224,083	41,751
2005	19,939,100	890,872	132,722	153,742	2,657,643	238,495	47,668
2006	20,579,925	911,561	133,443	172,844	2,797,500	250,313	51,355
2007	20,469,650	924,981	138,706	145,555	2,880,513	253,481	63,254
2008	19,106,850	804,959	123,949	107,583	2,232,238	186,655	48,767
2009	18,724,475	617,440	97,815	81,251	1,846,845	122,614	31,158
2010	18,421,425	569,523	92,327	75,488	1,659,525	100,542	25,941
2011	17,632,675	512,584	82,783	70,736	1,323,371	78,966	20,217
2012	17,139,000	408,537	68,988	61,578	1,112,233	51,327	13,249
2013	17,344,175	404,284	68,368	51,726	982,095	41,994	10,790
2014	17,866,050	424,625	71,411	52,255	991,202	43,043	7256
2015	18,341,550	458,023	76,941	53,099	1,059,440	48,813	8306
2016	18,824,825	489,065	81,216	58,207	1,095,710	53,579	8936
2017	19,327,725	515,082	86,593	60,259	1,150,639	61,375	10,500
2018	19,779,300	532,977	90,151	60,314	1,266,197	69,420	12,147
2019	19,773,600	562,756	96,777	51,445	1,340,185	73,666	14,828

shows data from 2000 to 2019 from the Spanish National Institute of Statistics (INE) on occupational accidents in Spain caused by accidents occurring at height put into the context of both the construction sector and all economic sectors (in general).

In absolute terms, if the data from the last 20 years are analysed, it can be deduced that, in relation to the employed population, which has risen by 22%, the number of accidents has fallen by 40% and the number of falls by 41%. However, in the construction sector, the number of building permits has fallen by 58%, while the number of people employed in the sector has fallen by 22%. Nevertheless, the number of accidents has fallen by 69% and the number of falls by 67%. In relative terms, as Figure 2 shows, a number of interesting conclusions can be drawn:

• The number of accidents per thousand people employed has fallen by 51%, while, in the construction sector, it has fallen by 61%. However, this number per thousand in the construction sector is still double the number of accidents per thousand overall;
• The number of falls per ten thousand people employed has fallen by 52%, while, in the construction sector, it has fallen by 58%. However, in the construction sector, this number is currently almost two and a half times that of the number per thousand overall. Note, also, that, over the last years of the real-estate bubble and subsequent crisis (2006–2012), this ratio tripled, confirming that, during those years, works were significantly less safe;
• The number of falls per thousand accidents has fallen by barely 2%, while, in the construction sector, it has risen by 8%. In addition, in the construction sector, this number is the same as the number per thousand overall, although, over the last years of the real-estate bubble and subsequent crisis (2006–2012), this ratio doubled.

Figure 2. Accidents caused by falls from height in the construction sector and in all economic sectors (in general). Source: INE.

Figure 2. Accidents caused by falls from height in the construction sector and in all economic sectors (in general). Source: INE.

Advances in the field of occupational safety and health stem mainly from legislative progress [57]. In Spain, the minimum requirements for safety and health on construction sites are regulated by Royal Decree 1627/1997 [58], the implementation decree of the general law on the Prevention of Occupational Risks 31/1995 and the transposition of European Directive 92/57/EEC on minimum safety and health requirements for construction sites. However, the existence of this regulatory framework has not brought about a proportional decrease in occupational accidents, as can be seen in Figure 2. This is due to the increase in construction works during the years of the real-estate bubble, as well as the reduction in investment in safety and health in the subsequent years of crisis. In addition, there has also not been adequate technical contributions to enable the correct application of the standards.

3.2. Description of Temporary Edge Protection Systems (TEPS)

This section describes the characteristics, manufacturing materials used, dimensions, versatility and approximate assembly times of the most common TEPS in the construction sector in Spain.

• Systems: There is a wide variety of collective protection systems available on the market for the construction sector, such as systems embedded in the concrete in plastic cartridges inserted into slabs, jaw-type posts tightened in concrete structures or fixed to metal profiles, clamp-type posts that can be fixed onto a wide range of slab edges, integrated safety systems used in the execution of formwork, telescopic vertical struts, V-type safety nets, etc. Likewise, within these systems, there are railings, nets, plinths and wire mesh which cover the gaps through which falls can occur, as well as anchoring systems, including flanges and uprights, with specific designs for their coupling;
• Materials: The most commonly used material for the manufacture of railings and uprights is galvanised steel [59], although other types of material such as aluminium [60], wood [61] or high-density polyethylene (HDPE) [62] can also be found, among others.

Of all the systems mentioned, this study focuses on the TEPS [63] most commonly used in construction, shown in Figure 3, since the elements they include (railings and uprights) are the most versatile and easy to find on the market and are also easier to assemble and disassemble than other systems.

If assembly times are taken into account, the selected TEPS are quick to install. First, prior to the concrete pour, the cartridge is inserted into the slab where the upright will be installed. Next, the upper and intermediate horizontal railings are coupled to the existing plates in the upright. However, due to their static geometry, often horizontal railings are bent to make a corner and, on other occasions, anchored with cables to the lengthwise limitation. On those occasions, assembly time is increased due to the lack of adaptability, which requires improvisation. On this last point, it must be noted that guard rails should not be used here as not only do they not prevent the fall of people from different levels but, sometimes, due to their poor construction and lack of strength, are the cause of the accident. A railing that is not strong enough is a real trap [65].

Based on the commercial solution mentioned above, Figure 4 shows the modelling of the elements of the selected TEPS:

3.3. Behaviour of the TEPS When Subjected to Different Loads and Materials

This section focuses on the response shown by a series of selected materials to the static load tests that must be met by the selected TEPS, classified as Class A by the UNE-EN 13374:2013+A1:2019 standard, as they are a means of collective protection used in construction works to prevent the fall of people and materials from horizontal surfaces or those with a slight inclination (up to 10 degrees) to a lower level, providing resistance to static loads, such as the weight of a worker leaning on the protection or resting his hand on it while walking alongside, as well as stopping a worker walking or falling in the direction of the protection. For this study, the materials of the two main elements that constitute the TEPS are selected (posts and railings), leaving the baseboards outside of the scope).

• Metals:

  ◦

  Galvanised steel (GS DX51D S280GD Z200), taken as a reference material;

  ◦

  Aluminium alloy 2024 (Al 2024 T3);

  ◦

  Cast iron (CI A536 80-55-06);

• Plastics:

  ◦

  Polyethylene terephthalate (PET);

  ◦

  Polyvinyl chloride (PVC);

• Composite materials:

  ◦

  Glass-fibre-reinforced polyester (GFRP);

  ◦

  Carbon-fibre-reinforced polymer (CFRP).

The characteristics of the chosen materials for this research are listed in

Table 2. Mechanical and strength characteristics of the materials used.

Material	GS	AL	CI	PET	PVC	GFRP	CFRP
Shear Modulus (N/mm2)	8.2 × 104	2.8 × 104	6.4 × 104	1.3 × 103	3.5 × 103	3.1 × 104	5.3 × 104
Young Modulus (N/mm2)	2.1 × 105	7.4 × 104	1.7 × 105	1.2 × 104	7.0 × 103	7.5 × 104	2.4 × 105
Density (g/cm3)	7.8	2.8	7.5	1.3	1.4	1.8	1.9
Shear Strength (N/mm2)	500	280	500	100	30	200	260
Yield Strength (N/mm2)	280	290	410	55	60	465	540
Tensile Strength (N/mm2)	360	440	600	150	110	1750	4170

.

In order to observe the behaviour of the different materials applied to the traditional Class A TEPS as a whole, the relevant calculations are carried out using the analytical method [66]. The tests carried out are the following, as summarised in Figure 5:

• Serviceability limit state (SLS):

  1.

  Elastic deflection requirement: The purpose is to evaluate the deformation capacity of the system (the system must be deformed but without exceeding a maximum limit) under a serviceability limit state criterion based on a horizontal characteristic point load perpendicular to the system of 300 N (F_T1) applied to the centre of the railing not exceeding 55 mm of deflection. Analytically, this condition is reflected in Equation (1):

  $${\mathsf{\delta }}_{\mathrm{ST}}=55 \ge \frac{{\mathrm{F}}_{\mathrm{T}1}\times {\mathrm{L}}_{\mathrm{G}}^3}{48\times \mathrm{E}\times {\mathrm{I}}_{\mathrm{G}}}+\frac{\frac{{\mathrm{F}}_{\mathrm{T}1}}{2}\times {\mathrm{L}}_{\mathrm{P}}^3}{3\times \mathrm{E}\times {\mathrm{I}}_{\mathrm{P}}}={\mathsf{\delta }}_{\mathrm{RT}}$$

  (1)

• Ultimate limit state (ULS):

  2.

  Flexural strength requirement: The purpose is to evaluate the bending capacity of the TEPS under the criterion of ultimate limit state (with increased loads) and determine the ultimate strength of the system from a horizontal point load perpendicular to the system of 300 N plus 50% (F_H1,d) applied to the centre of the railing. Analytically, this condition is reflected in Equation (2):

  $${\mathrm{M}}_{\mathrm{SD}}=\frac{{\mathrm{F}}_{\mathrm{H}1,\mathrm{d}}{\times \mathrm{L}}_{\mathrm{G}}}{4}\le \frac{{\mathrm{W}}_{\mathrm{G}}{\times \mathrm{f}}_{\mathrm{y}}}{{\mathsf{\gamma }}_{\mathrm{M}}}={\mathrm{M}}_{\mathrm{RD}}$$

  (2)

  3.

  Shear strength requirement: The purpose is to evaluate the shear capacity of the TEPS under the criterion of ultimate limit state (with increased loads) and determine the ultimate strength of the system from a horizontal point load perpendicular to the system of 300 N plus 50% (F_H1,d) applied to the centre of the railing. Analytically, this condition is reflected in Equation (3):

  $${\mathrm{V}}_{\mathrm{SD}}=\frac{{\mathrm{F}}_{\mathrm{H}1,\mathrm{d}}}{2}\le {\mathrm{A}}_{\mathrm{VG}}\times \frac{{\mathrm{f}}_{\mathrm{y}}/\sqrt{3}}{{\mathsf{\gamma }}_{\mathrm{M}}}={\mathrm{V}}_{\mathrm{RD}}$$

  (3)

• Ultimate limit state with parallel load (PL ULS):

  4.

  Parallel load bend resistance requirement: The purpose is to verify the bending behaviour of the system against parallel horizontal actions from a horizontal point load parallel to the system of 200 N plus 50% (F_H3,d) applied at the end of the post. Analytically, this condition is reflected in Equation (4):

  $${\mathrm{M}}_{\mathrm{SD}}{=\mathrm{F}}_{\mathrm{H}3,\mathrm{d}}{\times \mathrm{L}}_{\mathrm{P}}\le \frac{{\mathrm{W}}_{\mathrm{P}}{\times \mathrm{f}}_{\mathrm{y}}}{{\mathsf{\gamma }}_{\mathrm{M}}}={\mathrm{M}}_{\mathrm{RD}}$$

  (4)

  5.

  Parallel load shear resistance requirement: The purpose is to verify the shear behaviour of the system against parallel horizontal actions from a horizontal point load parallel to the system of 200 N plus 50% (F_H3,d) applied at the end of the post. Analytically, this condition is reflected in Equation (5):

  $${\mathrm{V}}_{\mathrm{SD}}{=\mathrm{F}}_{\mathrm{H}3,\mathrm{d}}\le {\mathrm{A}}_{\mathrm{VP}}\times \frac{{\mathrm{f}}_{\mathrm{y}}/\surd 3}{{\mathsf{\gamma }}_{\mathrm{M}}}={\mathrm{V}}_{\mathrm{RD}}$$

  (5)

• Ultimate limit state with accidental loads (AL ULS):

  6.

  Accidental load on elastic deflection requirement: The purpose is to evaluate the deformation (sag) of the system in front of a vertically descending point load of 1.25 kN (F_D) applied to the centre of the railing not exceeding 300 mm of deflection. Analytically, this condition is reflected in Equation (6):

  $${\mathsf{\delta }}_{\mathrm{SD}}=300 \ge \frac{{\mathrm{F}}_{\mathrm{D}}\times {\mathrm{L}}_{\mathrm{G}}^3}{48\times {\mathrm{f}}_{\mathrm{u}}\times {\mathrm{I}}_{\mathrm{G}}}={\mathsf{\delta }}_{\mathrm{RD}}$$

  (6)

  7.

  Accidental load bending requirement: The purpose is to evaluate the bending resistance to a vertically descending point load of 1.25 kN (F_D) applied to the centre of the railing. Analytically, this condition is reflected in Equation (7):

  $${\mathrm{M}}_{\mathrm{SD}}=\frac{{\mathrm{F}}_{\mathrm{D}}{\times \mathrm{L}}_{\mathrm{G}}}{4}\le {\mathrm{W}}_{\mathrm{G}}{\times \mathrm{f}}_{\mathrm{u}}={\mathrm{M}}_{\mathrm{RD}}$$

  (7)

  8.

  Accidental load shear requirement: The purpose is to evaluate the shear resistance of the system against a vertically descending point load of 1.25 kN (F_D) applied to the centre of the railing. Analytically, this condition is reflected in Equation (8):

  $${\mathrm{V}}_{\mathrm{SD}}=\frac{{\mathrm{F}}_{\mathrm{D}}}{2}\le {\mathrm{A}}_{\mathrm{VG}}\times \frac{{\mathrm{f}}_{\mathrm{u}}}{\surd 3}={\mathrm{V}}_{\mathrm{RD}}$$

  (8)

These analytical results are verified by finite-element analysis (FEA) [67] carried out with Autodesk Inventor software, with which the preliminary design is undertaken at a later stage. In the case of deflection tests, the results are direct. However, for the rest of the tests, the equivalent von Mises stress is provided by the software, which is expressed as shown in Equation (9):

$${\mathsf{\sigma }}_{\mathrm{VM}}=\sqrt{{\mathsf{\sigma }}_{\mathrm{x}}^2+3\times {\mathsf{\tau }}_{\mathrm{xz}}^2}\le \sqrt{{(\frac{{\mathrm{M}}_{\mathrm{y}}}{{\mathrm{W}}_{\mathrm{y}}})}^2+3\times {(\frac{{\mathrm{V}}_{\mathrm{y}}}{{\mathrm{A}}_{\mathrm{y}}})}^2}$$

(9)

where M_y is the bending moment (in N·m), W_y is the moment of resistance (in m³), V_y is the shear strength (in N) and A_y is the section area (in m²).

3.4. Classification of Attributes for the Preliminary Design of New TEPS

This section is responsible for analysing the attributes that must be incorporated into the new collective protection systems being designed, assessing basic, desired and motivating qualities through the application of the Kano model and its variants. Once the questionnaire is fully developed, it is distributed through the official colleges of technical experts with professional authority for the projection, direction, supervision and coordination of construction works in Spain (architects, technical architects, industrial engineers, industrial technical engineers, civil engineers, technical engineers of public works and civil engineers). Next, it is necessary to encode the answers given for each requirement according to the classification possibilities provided by the methodology, assigning the values of R (reverse requirement), Q (questionable requirement), A (attractive requirement), M (mandatory requirement), O (one dimensional) and I (indifferent), as shown in double entry

Table 3. Encoding responses according to Kano model.

Coding		Dysfunctional Requirements (Negative Questions)
		1	2	3	4	5
Functional requirements (positive questions)	5	C	A	A	A	O
	4	R	I	I	I	M
	3	R	I	I	I	M
	2	R	I	I	I	M
	1	R	R	R	R	C

. In functional questions, the scale goes from 1 (I dislike it a lot) to 5 (I like it a lot). Conversely, in dysfunctional questions, the scale goes from 1 (I like it a lot) to 5 (I dislike it a lot).

Based on the literature review [23,29,30,68,69,70,71,72,73,74] and information obtained in the field by the authors (in the design, supervision and coordination of their own works), an initial list of 42 requirements is elaborated and classified into 7 categories, as listed in

Table 4. Initial list of requirements.

Dimension	ID	Potential Requirement
Security	SE1	Resistance to mechanical stress
	SE2	Resistance to extreme temperatures
	SE3	Resistance to humidity
	SE4	Resistance to bad weather
	SE5	Safety through locking parts
	SE6	Technical support service
Ergonomics	ER1	Decomposition into low-weight elements
	ER2	Assembly/disassembly with ease
	ER3	Handling individually
	ER4	Functionality in different lengths and heights
	ER5	Configuration of a folding system
	ER6	Configuration of an extendable system
Sustainability	SO1	Use of fire-resistant materials
	SO2	Use of degradable materials
	SO3	Use of durable materials
	SO4	Use of coating materials
	SO5	Availability of spare parts
	SO6	Availability of repair service
Fabricability	FA1	Manufacturing with light materials
	FA2	Manufacturing with simple materials
	FA3	Manufacturing with composite materials
	FA4	Manufacturing with recycled materials
	FA5	Manufacturing with manual/artisanal materials
	FA6	Manufacturing with industrial/standardised materials
Efficiency	EF1	Transport by truck
	EF2	Transport by van
	EF3	Transport by car
	EF4	Storage in racks
	EF5	Storage in piles
	EF6	Procurement at low cost
Flexibility	FL1	Use of bright colours
	FL2	Use of neutral colours
	FL3	Use of luminescent materials
	FL4	Use of lighting elements
	FL5	Use of reflective elements
	FL6	Use of existing elements in the market
Adaptability	AD1	Manoeuvrability in different workplaces
	AD2	Compatibility with existing systems
	AD3	Versatility for use on different types of work sites
	AD4	Procurement of customised system elements
	AD5	Commercialisation free of charge
	AD6	Commercialisation through exclusive representation

.

Since the Kano model is used in a context where an eminently qualitative approach takes precedence, there is a statistical test that evaluates the significance of the classification carried out. To do this, the statistic E is used, obtained according to Equation (10) for the number of responses received (n), which is compared with the absolute difference (F) of the two frequencies (a and b) most voted of the alternatives (R, Q, A, M, O and I), as indicated in Equation 11, verifying that this difference is greater than this statistic [75].

$$\mathrm{E}=\sqrt{\frac{(\mathrm{a}+\mathrm{b})\times (2\mathrm{n}-\mathrm{a}-\mathrm{b})}{2\mathrm{n}}}$$

(10)

$$\mathrm{F}=\mathrm{a}-\mathrm{b}$$

(11)

Next, the requirements are classified by the increase or decrease in the satisfaction of potential users due to the inclusion or not of the requirement in the product, for which two other statistics are used, according to Equations (12) and (13). On the one hand, S is the perception of being better with its inclusion than without it and satisfying attractive and one-dimensional requirements. On the other hand, D is the perception of being worse without its inclusion than with it and not satisfying mandatory or one-dimensional requirements. Neither equation includes the reverse or questionable requirements precisely due to their confusing nature. S represents the ability of that attribute (potential requirement) to increase user satisfaction, while D represents the ability to reduce user dissatisfaction. High values of S and D give one-dimensional requirements (requirements that are wanted and must be incorporated). High values of S and low values of D give attractive requirements (requirements we wish to incorporate: exciting attributes). Low values of S and high values of D give mandatory requirements (requirements that must be incorporated: basic attributes). Low values of S and D result in indifferent requirements.

$$\mathrm{S}=\frac{(\mathrm{A}+\mathrm{O})}{(\mathrm{A}+\mathrm{O}+\mathrm{M}+\mathrm{I})}$$

(12)

$$\mathrm{D}=\frac{(\mathrm{U}+\mathrm{M})}{- (\mathrm{A}+\mathrm{O}+\mathrm{M}+\mathrm{I})}$$

(13)

Nevertheless, some aspects of analysing the needs of potential users are not completely resolved. If the Kano model asks about very general functions, interviewees have a concrete opinion. However, if the questionnaire asks about very specific functions, the responses of most respondents lead to an indifferent requirement. Consequently, extremely detailed questions can increase the noise level to a point where the requirements are indifferent. One way to modify the statistics is to calculate the vector (X, Y) for each requirement (i) according to the scores referred to in

Table 5. Scores of functional and dysfunctional responses.

Responses	Functional Form of Question		Dysfunctional Form of Question
	Scale	Score	Scale	Score
I like it that way	5	+1.00	1	−0.50
It must be that way	4	+0.50	2	−0.25
I am neutral	3	±0.00	3	±0.00
I can live with it that way	2	−0.25	4	+0.50
I dislike it that way	1	−0.50	5	+1.00

and Equations (14) and (15). Therefore, dimension X indicates dissatisfaction if a requirement is not included and Y satisfaction if it is [76].

$${\mathrm{X}}_{\mathrm{i}}=\frac{1}{\mathrm{n}} {\displaystyle \sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{x}}_{\mathrm{ij}}$$

(14)

$${\mathrm{Y}}_{\mathrm{i}}=\frac{1}{\mathrm{n}} {\displaystyle \sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{y}}_{\mathrm{ij}}$$

(15)

The classification of the revised Kano model shows how a potential user classifies a possible requirement of a product by comparing the satisfaction or dissatisfaction of its inclusion against the better or worse performance of that product [77]. However, if the questionnaire additionally asks about the normalised importance (G) given by potential users to that requirement, then the weighted vector (C1, C2) [78] can be determined, taking into account the intrinsic importance of each attribute. This provides a quantifiable statistical result which can be fed and integrated into an analytical model, such as a quality function deployment (QFD) model [79], that can be used in more advanced stages [39].

3.5. Hierarchisation of Alternative Materials for the Preliminary Design of New TEPS

This section is responsible for hierarchising the alternative materials that can be used for the new collective protection equipment to be designed after carrying out the appropriate tests according to the attributes considered as potential requirements (mandatory, attractive and one dimensional) resulting from the Kano models. To do this, the AHP method is used, taking as decision makers a panel of independent research experts thanks to the collaboration of the professional associations whose associate members have explicit legal authority in the construction process and to the Labour and Social Security Inspectorate (ITSS). The panel of experts is established following stakeholder theory [80], taking into account business owners, architecture and engineering firms, contractor corporations, public inspection bodies and occupational mutual insurance companies (OMIC) [81]. This panel needs to have the necessary knowledge and experience to validate (or add and/or delete) the proposed scheme of criteria (dimensions), sub-criteria (attributes) and alternatives (materials) and prioritise each of these categories in pairs, as well as decide the comparison scales, either the Saaty scale [80], compiled in

Table 6. Saaty scale for AHP.

Intensity of Importance	Definition	Explanation
1	Equal importance	Equal contribution to the objective
3	Moderate importance	Experience and judgement slightly favour one over another
5	Strong importance	Experience and judgement strongly favour one over another
7	Very strong importance	One is favoured very strongly over another
9	Extreme importance	The evidence favouring one over another is the highest
2, 4, 6, 8	Intermediate values	Judgements between defined prior intensities
1/2, 1/3, 1/4, 1/5, 1/6, 1/7, 1/8, 1/9	Reciprocal values	Opposite judgements concerning defined prior intensities

, or a natural scale. Valuations are made by taking a consensus value.

To check the consistency of the valuations, the consistency ratio (CR) is used, according to Equation (16), which, for each matrix obtained, must be less than 5% for matrices of order 3, 8% for matrices of order 4 and 10% for matrices of order 5 or higher [82].

$$\mathrm{CR}=\frac{({\mathsf{\lambda }}_{\max }-\mathrm{n})/\left(\mathrm{n}-1\right)}{\mathrm{RI}}$$

(16)

where λ_max is the primary eigenvalue of each comparison matrix, n is the order of each matrix and RI is the random index [83], the values of which are listed in

Table 7. Index ratio values based on the order of the comparison matrix.

n	1	2	3	4	5	6	7	8	9	10
RI	0	0	0.525	0.882	1.115	1.252	1.341	1.404	1.452	1.484

.

The pairwise comparisons of dimensions, attributes and alternative materials are generated and organised into square matrices, as shown in Equation (17):

$${\mathrm{A}}_{\mathrm{w}}=\left[\begin{array}{ccccc}{\mathrm{a}}_{11}& \cdots & {\mathrm{a}}_{1\mathrm{i}}& \cdots & {\mathrm{a}}_{1\mathrm{n}}\\ \cdots & \cdots & \cdots & \cdots & \cdots \\ {\mathrm{a}}_{\mathrm{j}1}& \cdots & {\mathrm{a}}_{\mathrm{ji}}& \cdots & {\mathrm{a}}_{\mathrm{jn}}\\ \cdots & \cdots & \cdots & \cdots & \cdots \\ {\mathrm{a}}_{\mathrm{n}1}& \cdots & {\mathrm{a}}_{\mathrm{ni}}& \cdots & {\mathrm{a}}_{\mathrm{nn}}\end{array}\right]$$

(17)

where a_ij is the comparison between item i and j considering that a₁₁ = a_nn because of homogeneity, a_ij × a_ji = 1 because of reciprocity and a_ik × r_kj = r_ij because of transitivity. The criteria weights of each item to be considered are normalised by mathematically solving for a non-zero eigenvalue, as described in Equation (18):

$${\displaystyle \sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{a}}_{\mathrm{ij}}{\mathrm{w}}_{\mathrm{j}}={\mathsf{\lambda }}_{\max }{\mathrm{w}}_{\mathrm{i}}$$

(18)

where w is the criteria weights (eigenvectors), and λ_max is the maximum eigenvalue of the model. On the other hand, the eigenvector illustrates the relative weights of each criterion under every level of the model. Then, the priority weights of each level can be aggregated using the geometric mean method. Answers from each respondent are agreed and then synthesised into a single priority vector in order to obtain an overall computation of the priorities for each factor in the model according to Equation (19):

$$\mathrm{G}=\left[\begin{array}{c}{(1\times \cdots \times {\mathrm{a}}_{1\mathrm{i}}\times \cdots \times {\mathrm{a}}_{1\mathrm{n}})}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$\mathrm{n}$}\right.}\\ \cdots \\ {({\mathrm{a}}_{\mathrm{j}1}\times \cdots \times {\mathrm{a}}_{\mathrm{ji}}\times \cdots \times {\mathrm{a}}_{\mathrm{jn}})}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$\mathrm{n}$}\right.}\\ \cdots \\ {({\mathrm{a}}_{\mathrm{n}1}\times \cdots \times {\mathrm{a}}_{\mathrm{ni}}\times \cdots \times 1)}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$\mathrm{n}$}\right.}\end{array}\right]$$

(19)

where G is the geometric mean of each factor in the hierarchy, a is the weight provided by the experts panel and n is the order of each pairwise comparison matrix. Finally, the global priority weight of each parameter is computed using Equation (20):

$$\mathrm{G}{\mathrm{W}}_{\mathrm{pi}}={\mathrm{W}}_{\mathrm{fi}}\times {\mathrm{W}}_{\mathrm{ci}}$$

(20)

where i is the hierarchy level, W_f is the factor local priority weightage and W_C is the category local priority weightage.

4. Results

4.1. Results Compared with Static Load SLS Requirements

Table 8. SLS results of maximum deflection at the midpoint of the upper horizontal railing.

Test 1	GS	AL	CI	PET	PVC	GFRP	CFRP
Maximum Deflection (mm)	55	55	55	55	55	55	55
Analytic Deflection (mm)	26	74	33	454	779	73	23
FEA Deflection (mm)	21	60	26	365	628	59	19
Compliance?	OK	KO	OK	KO	KO	KO	OK

shows the results of this test. From the results obtained after the static structural analysis, applying a horizontal point load to the midpoint of the upper horizontal railing, the metals show optimal behaviour, obtaining a deflection lower than the maximum allowed by the UNE-EN 13374:2013+A1:2019 standard [64], except for the AL, which requires a new sizing. In the case of the composite materials, the CFRP amply fulfils requirements, but the GFRP requires new sizing. In the case of plastics, the requirement is not met.

Likewise, Figure 6 shows the deformation of the reference material (GS) obtained in the FEA in test 1 compared to the static load SLS requirements. The rest of the materials (alternative materials) are shown in Figure A1 of Appendix A.

4.2. Results Compared with Static Load ULS Requirements

Table 9. ULS results of bending moment at the midpoint of the upper horizontal railing.

Test 2	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Bending Moment (N·m)	281	281	281	281	281	281	281
Analytic Bending Moment (N·m)	555	575	813	109	119	922	1071
Max. von Mises Stress (N/mm2)	342	354	500	67	73	568	659
FEA von Mises Stress (N/mm2)	331	332	330	356	345	338	341
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the flexural results of this test. From the results obtained after the static structural analysis, applying a horizontal point load to the midpoint of the upper horizontal railing, both composite materials and metals show optimal behaviour, obtaining a bending resistance higher than the minimum required by the UNE-EN 13374:2013+A1:2019 standard. In the case of plastics, the requirement is not met.

On the other hand,

Table 10. ULS results of shear stress at the midpoint of the upper horizontal railing.

Test 3	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Shear Strength (N)	225	225	225	225	225	225	225
Analytic Shear Strength (N)	30,472	31,560	44,619	5986	6530	50,605	58,767
Max. von Mises Stress (N/mm2)	342	354	500	67	73	568	659
FEA von Mises Stress (N/mm2)	331	332	330	356	345	338	341
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the shearing results of this test. From the results obtained after the static structural analysis, applying a horizontal point load to the midpoint of the upper horizontal railing, all materials show optimal behaviour, since, in all cases, a shear resistance higher than the minimum required by the UNE-EN 13374:2013+A1:2019 standard is obtained. However, according to the FEA analysis, a detailed study of the railing–post coupling element is required, as in the previous test.

Likewise, Figure 7 shows the von Mises stresses of the reference material (GS) obtained in the FEA in tests 2 and 3 compared to the ULS requirements of static load. The remaining materials (alternative materials) are shown in Figure A2 of Appendix A.

4.3. Results Compared with Parallel Load ULS Requirements

On the one hand,

Table 11. PL ULS results of bending moment at the base of the upright.

Test 4	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Bending Moment (N·m)	360	360	360	360	360	360	360
Analytic Bending Moment (N·m)	736	762	1077	145	158	1222	1419
Max. von Mises Stress (N/mm2)	342	354	500	67	73	568	659
FEA von Mises Stress (N/mm2)	104	91	103	86	87	87	87
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the bending results of this test. From the results obtained after the static structural analysis, applying a parallel point load to the upper end of the upright, both composite materials and metals show optimal behaviour, obtaining a bending resistance at the base of the upright higher than the minimum required by the UNE-EN 13374:2013+A1:2019 standard. On the contrary, in the case of plastics, the requirement is not met.

On the other hand,

Table 12. PL ULS results of shear stress at the base of the upright.

Test 5	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Shear Strength (N)	300	300	300	300	300	300	300
Analytic Shear Strength (N)	35,089	36,342	51,380	6892	7519	58,272	67,671
Max. von Mises Stress (N/mm2)	342	354	500	67	73	568	659
FEA von Mises Stress (N/mm2)	104	91	103	86	87	87	87
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the results of this test. From the results obtained after the static structural analysis, applying a parallel point load to the upper end of the upright, all materials show optimal behaviour, since, in all cases, a shear resistance higher than the minimum required by the UNE-EN 13374:2013+A1:2019 standard is obtained. However, according to the FEA analysis, a detailed study of the railing–post coupling element is required, as in the previous test.

Likewise, Figure 8 shows the von Mises stresses of the reference material (GS) obtained in the FEA in tests 4 and 5 compared to the ULS parallel load requirements. The rest of the materials (alternative materials) are shown in Figure A3 of Appendix A.

4.4. Results Compared with Accidental Static Load ULS Requirements

Table 13. AL ULS results of maximum deflection at the midpoint of the top horizontal railing.

Test 6	GS	AL	CI	PET	PVC	GFRP	CFRP
Maximum Deflection (mm)	300	300	300	300	300	300	300
Analytic Deflection (mm)	69	195	85	1197	2052	192	60
FEA Deflection (mm)	19	53	23	319	548	52	17
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the bending results of this test. From the results obtained after the static structural analysis, applying an accidental point load to the midpoint of the upper horizontal railing, both composite materials, including GFRP and metals, show optimal behaviour, obtaining a deflection lower than the maximum allowed by the UNE-EN 13374:2013+A1:2019 standard [64]. In the case of plastics, the requirement is not met.

Likewise, Figure 9 shows the deformation of the reference material obtained in the FEA in test 6 compared with the ELU requirements for accidental static load. The remaining materials (alternative materials) are shown in Figure A4 of Appendix A. It should be noted that there is a difference in results according to the analytical method compared to the FEA (almost four times lower). This is due to the design of the coupling between the railing and the post of the commercial model. The analytical method only takes into account the circular section of the elements. However, when a horizontal load is applied to the railing, it has a lower inertia than when a vertical load is applied, so the deflection is reduced.

On the other hand,

Table 14. AL ULS results of bending (breakage) at the midpoint of the upper horizontal railing.

Test 7	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Bending Moment (N·m)	781	781	781	781	781	781	781
Analytic Bending Moment (N·m)	785	959	1308	327	240	3816	9093
Max. von Mises Stress (N/mm2)	363	443	605	151	111	1766	4208
FEA von Mises Stress (N/mm2)	371	356	371	343	341	346	344
Compliance?	OK	OK	OK	KO	KO	OK	OK

shows the flexion results of this test. In this test, the UNE-EN 13374:2013+A1:2019 standard allows compliance with this requirement to be achieved until its breaking capacity is exhausted. From the results obtained after the static structural analysis, applying an accidental point load to the midpoint of the upper horizontal railing, on the part of the metals, both the reference material and the cast iron manage to meet the requirement. However, aluminium requires a new sizing. On the contrary, composite materials more than meet the requirement. In the case of plastics, the requirement is not met. It should also be noted that the stresses obtained for the reference material in the FEA analysis also exceed the established limit (by 2%), so this detail should be studied in the detail design if the reference material is selected and this connection is used.

In addition,

Table 15. Results ULS accidental load to shear stress at the midpoint of the upper horizontal railing.

Test 8	GS	AL	CI	PET	PVC	GFRP	CFRP
Min. Shear Strength (N)	625	625	625	625	625	625	625
Analytic Shear Strength (N)	33,519	33,519	49,081	6584	3591	23,942	31,125
Max. von Mises Stress (N/mm2)	363	443	605	151	111	1766	4208
FEA von Mises Stress (N/mm2)	371	356	371	343	341	346	344
Compliance?	KO	OK	OK	KO	KO	OK	OK

shows the results of this test. From the results obtained after the static structural analysis, applying an accidental point load to the midpoint of the upper horizontal railing, all materials show optimal behaviour, since, in all cases, a shear resistance higher than the minimum required by the UNE-EN 13374:2013+A1:2019 standard is obtained. However, in accordance with the FEA analysis, a detailed study of the railing–post coupling element is required, as in the previous test, for both the plastics and the reference material.

Finally, Figure 10 shows the von Mises stresses of the reference material obtained in the FEA in tests 7 and 8 compared to the ULS requirements of accidental static load. The remaining materials (alternative materials) are shown in Figure A5 of Appendix A.

4.5. Requirement Classification Results

After the end of the questionnaire submission period, 190 replies are received. In order to describe the sample set, a series of control questions is asked related to personal issues (gender, age and level of education), to the company people work for (size and position) and to project management (specific training, professional certification and experience).

Table 16. Population control questions for sample classification.

Control Questions	Responses	Units	Percentage
Age	<25 years	2	1.05%
	25–30 years	13	6.84%
	31–45 years	123	64.74%
	46–60 years	44	23.16%
	>60 years	8	4.21%
Organisational size	Self-employed	39	20.53%
	Micro enterprise	54	28.42%
	Small enterprise	31	16.32%
	Medium enterprise	23	12.11%
	Large enterprise	43	22.63%
Works complexity	Very simple	17	8.95%
	Simple	22	11.58%
	Normal	63	33.16%
	Complex	46	24.21%
	Very complex	42	22.11%

summarises the characteristics of the population sample. Most of the sample is in the middle part of their professional life, working in SME-type companies or self-employed in works of medium-to-large complexity.

The second part of the survey is shown in

Table A1. Functional scores, dysfunctional scores and importance of requirements collection.

Functional Scores					Dysfunctional Scores					Grade of Importance
1	2	3	4	5	1	2	3	4	5	0	1	2	3	4	5	6	7	8	9	10
		21	56	113		6	9	51	124					4	6	11	24	36	76	33
		83	85	22			28	44	118						7	33	36	84	24	6
		92	59	39			64	97	29					14	32	56	67	21
		64	97	29			11	33	146					3	41	64	46	32	4
	11	22	28	129		21	49	94	26						6	11	34	56	66	17
	44	67	23	56			31	133	26			6	9	14	26	67	51	13	4
		9	80	101			6	74	110								57	77	49	7
			43	147				26	164									9	44	137
8	17	61	66	38	24	32	31	34	69					14	21	34	64	46	11
		16	36	138			26	62	102							7	28	44	62	49
43	57	26	42	22	33	39	23	56	39				33	71	46	27	13
4	17	44	72	53		16	59	111	4				6	8	18	32	51	44	22	9
		72	89	29			17	46	127							26	79	74	11
21	36	54	26	53	24	57	41	24	44			11	23	44	69	23	14	6
			56	134			16	31	143							24	31	36	53	46
	27	62	72	29	44	41	49	37	19					24	71	81	14
		11	31	148		24	46	76	44					4	19	84	57	14	12
	17	28	98	47			54	83	53					11	63	76	34	6
		21	38	131			86	66	38						14	46	51	56	19	6
9	83	46	31	21		6	56	87	41						9	33	57	64	27
	6	11	34	139	4	11	59	82	34					11	33	74	41	22	9
		26	41	123			71	96	23						14	79	51	42	4
13	77	47	41	12	24	61	52	34	19				8	27	87	51	17
6	12	43	107	22		11	16	37	126					16	31	33	49	36	19	6
14	36	44	49	47			6	37	147						26	76	51	34	3
			93	97			7	49	134					14	24	36	44	41	19	12
		13	43	134			58	114	18						17	26	41	54	29	27
	7	18	62	103			12	74	104						6	33	72	53	26
	81	73	22	14	11	24	76	43	36			4	21	49	56	34	22	4
		66	83	31			14	55	121						4	16	79	44	31	16
8	26	59	54	43	9	13	29	76	63				6	24	49	36	31	26	18
11	14	31	76	58	12	23	59	52	44				21	29	47	61	24	8
6	43	101	24	16	14	21	91	46	18				24	44	71	34	11	6
	24	113	31	22		6	121	49	14			8	33	38	41	36	23	11
	43	57	29	61		56	66	39	29					7	31	39	61	31	21
	13	21	44	112		4	21	62	103					3	9	22	42	61	46	7
		13	54	123			7	26	157						2	36	26	41	47	38
		4	54	132			14	87	109								4	29	94	63
		17	69	104			21	63	106						4	13	22	31	84	36
		23	64	103			14	72	104					4	21	28	49	61	27
		47	103	40	2	7	41	94	36					2	11	89	46	29	13
44	25	51	34	36	13	21	49	63	44				8	45	66	56	15

of Appendix B, which shows the scores of 1−5 for functional and dysfunctional questions, as well as the degree of importance given to each attribute. Using this information,

Table A2. Functional/dysfunctional classification of requirements.

ID	Ri	Ci	Ai	Oi	Ui	Ii	Qi	Fi	Fi ≥ Qi	Si	Di	Type	Xi	Yi	Type	Gi	C1i	C2i	ri	αi	Type
SE1				11	113	66	16.06	47	Ok	0.59	−0.65	U	0.74	0.78	U	8.33	0.62	0.65	1.08	0.81 (46°)	U
SE2				96	22	72	15.97	24	Ok	0.12	−0.62	O	0.34	0.74	O	7.54	0.26	0.56	0.81	1.14 (65°)	O
SE3			10		29	151	16.06	122	Ok	0.21	−0.15	I	0.36	0.41	I	6.26	0.23	0.26	0.54	0.85 (49°)	I
SE4			0	117	29	44	15.89	73	Ok	0.15	−0.77	O	0.41	0.86	O	6.39	0.26	0.55	0.95	1.13 (65°)	O
SE5			103		26	61	15.93	42	Ok	0.68	−0.14	A	0.74	0.36	A	8.14	0.60	0.29	0.82	0.45 (26°)	A
SE6			30		26	134	15.93	104	Ok	0.29	−0.14	I	0.30	0.49	I	5.92	0.18	0.29	0.57	1.02 (59°)	I
ER1				9	101	80	16.06	21	Ok	0.53	−0.58	U	0.74	0.77	U	8.03	0.60	0.62	1.07	0.81 (46°)	U
ER2				17	147	26	16.02	121	Ok	0.77	−0.86	U	0.89	0.93	U	9.67	0.86	0.90	1.29	0.81 (46°)	U
ER3	24			31	38	97	15.39	59	Ok	0.23	−0.42	I	0.33	0.35	I	6.74	0.22	0.23	0.48	0.81 (46°)	I
ER4			36		102	52	15.79	50	Ok	0.73	−0.54	U	0.82	0.70	U	8.62	0.71	0.60	1.08	0.71 (40°)	U
ER5	43			17	22	108	15.74	65	Ok	0.15	−0.27	I	0.04	0.21	I	4.56	0.02	0.10	0.22	1.39 (80°)	I
ER6	4		49		4	133	16.07	84	Ok	0.28	−0.02	I	0.44	0.29	I	6.99	0.30	0.20	0.52	0.59 (34°)	I
SU1				98	29	63	15.89	35	Ok	0.15	−0.67	O	0.39	0.79	O	7.37	0.29	0.58	0.88	1.12 (64°)	O
SU2	24		9		44	113	15.84	69	Ok	0.32	−0.27	I	0.24	0.16	I	4.72	0.12	0.07	0.29	0.57 (33°)	I
SU3	0			9	134	47	16.06	87	Ok	0.71	−0.75	U	0.85	0.83	U	8.35	0.71	0.70	1.19	0.77 (44°)	U
SU4	44		10		19	117	15.89	73	Ok	0.20	−0.13	I	0.31	0.03	I	5.45	0.17	0.02	0.31	0.09 (5°)	I
SU5			104		44	42	15.68	60	Ok	0.78	−0.23	A	0.86	0.40	A	6.49	0.56	0.26	0.95	0.44 (25°)	A
SU6				6	47	137	16.07	90	Ok	0.25	−0.28	I	0.48	0.50	I	5.79	0.28	0.29	0.69	0.80 (46°)	I
FA1			93		38	59	15.76	34	Ok	0.69	−0.20	A	0.79	0.37	A	7.27	0.57	0.27	0.87	0.44 (25°)	A
FA2	9			20	21	140	15.89	119	Ok	0.12	−0.23	I	0.06	0.44	I	7.35	0.04	0.32	0.44	1.44 (82°)	I
FA3	4		105		34	47	15.76	58	Ok	0.75	−0.18	A	0.81	0.37	A	6.30	0.51	0.23	0.89	0.43 (24°)	A
FA4			100		23	67	15.96	33	Ok	0.65	−0.12	A	0.76	0.37	A	6.70	0.51	0.25	0.84	0.46 (26°)	A
FA5	24			7	12	147	16.00	123	Ok	0.07	−0.11	I	0.04	0.05	I	5.22	0.02	0.02	0.06	0.91 (52°)	I
FA6	6			104	22	58	15.91	46	Ok	0.12	−0.68	O	0.37	0.75	O	6.73	0.25	0.50	0.83	1.11 (64°)	O
EF1	14			100	47	29	15.66	53	Ok	0.27	−0.84	O	0.29	0.87	O	6.54	0.19	0.57	0.92	1.25 (71°)	O
EF2				37	97	56	15.77	41	Ok	0.51	−0.71	U	0.76	0.83	U	6.94	0.52	0.58	1.13	0.84 (48°)	U
EF3			116		18	56	16.01	60	Ok	0.71	−0.09	A	0.82	0.39	A	7.85	0.64	0.31	0.91	0.45 (26°)	A
EF4	0			1	103	86	16.08	17	Ok	0.54	−0.55	U	0.70	0.74	U	7.32	0.51	0.54	1.02	0.82 (47°)	U
EF5	11			22	14	143	15.94	121	Ok	0.08	−0.20	I	0.03	0.24	I	4.93	0.01	0.12	0.24	1.47 (84°)	I
EF6				90	31	69	15.87	21	Ok	0.16	−0.64	O	0.38	0.78	O	7.68	0.29	0.60	0.87	1.12 (64°)	O
FL1	9			20	43	118	15.89	75	Ok	0.24	−0.35	I	0.31	0.49	I	6.12	0.19	0.30	0.58	1.00 (57°)	I
FL2	12		14		44	120	15.93	76	Ok	0.33	−0.25	I	0.46	0.31	I	5.33	0.24	0.16	0.55	0.59 (34°)	I
FL3	14			2	16	158	16.03	142	Ok	0.09	−0.10	I	0.08	0.15	I	4.91	0.04	0.07	0.17	1.11 (64°)	I
FL4			8		14	168	16.07	154	Ok	0.12	−0.07	I	0.17	0.19	I	4.93	0.08	0.10	0.26	0.87 (50°)	I
FL5			32		29	129	15.89	97	Ok	0.32	−0.15	I	0.34	0.18	I	6.74	0.23	0.12	0.39	0.49 (28°)	I
FL6			9		103	78	16.06	25	Ok	0.59	−0.54	U	0.69	0.70	U	7.66	0.53	0.54	0.98	0.79 (45°)	U
AD1				34	123	33	15.84	89	Ok	0.65	−0.83	U	0.79	0.89	U	8.10	0.64	0.72	1.19	0.85 (49°)	U
AD2			23	0	109	58	15.96	51	Ok	0.69	−0.57	U	0.84	0.80	U	9.14	0.76	0.73	1.16	0.76 (44°)	U
AD3				2	104	84	16.08	20	Ok	0.55	−0.56	U	0.73	0.72	U	8.51	0.62	0.62	1.03	0.78 (45°)	U
AD4				1	103	86	16.08	17	Ok	0.54	−0.55	U	0.71	0.74	U	7.17	0.51	0.53	1.02	0.8 (46°)	U
AD5	2		4		36	148	16.07	112	Ok	0.21	−0.19	I	0.48	0.42	I	6.67	0.32	0.28	0.64	0.72 (41°)	I
AD6	44			8	36	102	15.65	58	Ok	0.25	−0.30	I	0.13	0.34	I	5.13	0.07	0.17	0.36	1.20 (69°)	I

of Appendix B is elaborated, using the Kano method to establish the classification of the requirements into indifferent, mandatory, one dimensional and attractive. A summary of this can be seen in Table 17. First, the number of respondents that classifies each requirement as reverse (R), questionable (C), mandatory (O), attractive (A) and one dimensional (U) is counted. Next, the significance of the answers is checked, for which the difference F must be greater than the Q statistic. Once this check is performed, the vector (S, D) is represented in a graph as the requirements help to increase satisfaction and/or decrease dissatisfaction (Figure 11, above). Next, the vector (X, Y) is calculated, which weighs dissatisfaction for potential users if a requirement is not included and satisfaction if it is (Figure 11, bottom left). Subsequently, the normalised importance G given by potential users to that requirement is measured to determine the weighted vector (C1, C2), including its polar coordinates (r, α) (Figure 11, bottom right). According to the data obtained, and ignoring indifferent requirements (I), the remaining requirements are classified into:

• Mandatory (M):
  SE2, SE4, SO1, FA6, EF1, EF6;
• Attractive (A):
  SE5, SO5, FA1, FA3, FA4, EF3;
• One dimensional (O):
  SE1, ER1, ER2, ER4, SO3, EF2, EF4, FL6, AD1, AD2, AD3, AD4.

Table 17. Classification of mandatory, attractive and one-dimensional requirements.

ID	Attribute	Si	Di	Type	Xi	Yi	Type	Gi	C1i	C2i	Type
SE	Security:
SE1	Resistance to mechanical stress	0.59	−0.65	U	0.74	0.78	U	8.33	0.62	0.65	U
SE2	Resistance to extreme temperatures	0.12	−0.62	O	0.34	0.74	O	7.54	0.26	0.56	O
SE4	Resistance to bad weather	0.15	−0.77	O	0.41	0.86	O	6.39	0.26	0.55	O
SE5	Safety through locking parts	0.68	−0.14	A	0.74	0.36	A	8.14	0.60	0.29	A
ER	Ergonomics:
ER1	Decomposition into low-weight elements	0.53	−0.58	U	0.74	0.77	U	8.03	0.60	0.62	U
ER2	Assembly/disassembly with ease	0.77	−0.86	U	0.89	0.93	U	9.67	0.86	0.90	U
ER4	Functionality in different lengths and heights	0.73	−0.54	U	0.82	0.70	U	8.62	0.71	0.60	U
SU	Sustainability:
SU1	Use of fire-resistant materials	0.15	−0.67	O	0.39	0.79	O	7.37	0.29	0.58	O
SU3	Use of durable materials	0.71	−0.75	U	0.85	0.83	U	8.35	0.71	0.70	U
SU5	Availability of spare parts	0.78	−0.23	A	0.86	0.40	A	6.49	0.56	0.26	A
FA	Fabricability:
FA1	Manufacturing with light materials	0.69	−0.20	A	0.79	0.37	A	7.27	0.57	0.27	A
FA3	Manufacturing with composite materials	0.75	−0.18	A	0.81	0.37	A	6.30	0.51	0.23	A
FA4	Manufacturing with recycled materials	0.65	−0.12	A	0.76	0.37	A	6.70	0.51	0.25	A
FA6	Manufacturing with industrial/standardised materials	0.12	−0.68	O	0.37	0.75	O	6.73	0.25	0.50	O
EF	Efficiency:
EF1	Transport by truck	0.27	−0.84	O	0.29	0.87	O	6.54	0.19	0.57	O
EF2	Transport by van	0.51	−0.71	U	0.76	0.83	U	6.94	0.52	0.58	U
EF3	Transport by car	0.71	−0.09	A	0.82	0.39	A	7.85	0.64	0.31	A
EF4	Storage in racks	0.54	−0.55	U	0.70	0.74	U	7.32	0.51	0.54	U
EF6	Procurement at low cost	0.16	−0.64	O	0.38	0.78	O	7.68	0.29	0.60	O
FL	Flexibility:
FL6	Use of existing elements in the market	0.59	−0.54	U	0.69	0.70	U	7.66	0.53	0.54	U
AD	Adaptability:
AD1	Manoeuvrability in different workplaces	0.65	−0.83	U	0.79	0.89	U	8.10	0.64	0.72	U
AD2	Compatibility with existing systems	0.69	−0.57	U	0.84	0.80	U	9.14	0.76	0.73	U
AD3	Versatility for use on different types of work sites	0.55	−0.56	U	0.73	0.72	U	8.51	0.62	0.62	U
AD4	Procurement of customised system elements	0.54	−0.55	U	0.71	0.74	U	7.17	0.51	0.53	U

Table 17. Classification of mandatory, attractive and one-dimensional requirements.

ID	Attribute	Si	Di	Type	Xi	Yi	Type	Gi	C1i	C2i	Type
SE	Security:
SE1	Resistance to mechanical stress	0.59	−0.65	U	0.74	0.78	U	8.33	0.62	0.65	U
SE2	Resistance to extreme temperatures	0.12	−0.62	O	0.34	0.74	O	7.54	0.26	0.56	O
SE4	Resistance to bad weather	0.15	−0.77	O	0.41	0.86	O	6.39	0.26	0.55	O
SE5	Safety through locking parts	0.68	−0.14	A	0.74	0.36	A	8.14	0.60	0.29	A
ER	Ergonomics:
ER1	Decomposition into low-weight elements	0.53	−0.58	U	0.74	0.77	U	8.03	0.60	0.62	U
ER2	Assembly/disassembly with ease	0.77	−0.86	U	0.89	0.93	U	9.67	0.86	0.90	U
ER4	Functionality in different lengths and heights	0.73	−0.54	U	0.82	0.70	U	8.62	0.71	0.60	U
SU	Sustainability:
SU1	Use of fire-resistant materials	0.15	−0.67	O	0.39	0.79	O	7.37	0.29	0.58	O
SU3	Use of durable materials	0.71	−0.75	U	0.85	0.83	U	8.35	0.71	0.70	U
SU5	Availability of spare parts	0.78	−0.23	A	0.86	0.40	A	6.49	0.56	0.26	A
FA	Fabricability:
FA1	Manufacturing with light materials	0.69	−0.20	A	0.79	0.37	A	7.27	0.57	0.27	A
FA3	Manufacturing with composite materials	0.75	−0.18	A	0.81	0.37	A	6.30	0.51	0.23	A
FA4	Manufacturing with recycled materials	0.65	−0.12	A	0.76	0.37	A	6.70	0.51	0.25	A
FA6	Manufacturing with industrial/standardised materials	0.12	−0.68	O	0.37	0.75	O	6.73	0.25	0.50	O
EF	Efficiency:
EF1	Transport by truck	0.27	−0.84	O	0.29	0.87	O	6.54	0.19	0.57	O
EF2	Transport by van	0.51	−0.71	U	0.76	0.83	U	6.94	0.52	0.58	U
EF3	Transport by car	0.71	−0.09	A	0.82	0.39	A	7.85	0.64	0.31	A
EF4	Storage in racks	0.54	−0.55	U	0.70	0.74	U	7.32	0.51	0.54	U
EF6	Procurement at low cost	0.16	−0.64	O	0.38	0.78	O	7.68	0.29	0.60	O
FL	Flexibility:
FL6	Use of existing elements in the market	0.59	−0.54	U	0.69	0.70	U	7.66	0.53	0.54	U
AD	Adaptability:
AD1	Manoeuvrability in different workplaces	0.65	−0.83	U	0.79	0.89	U	8.10	0.64	0.72	U
AD2	Compatibility with existing systems	0.69	−0.57	U	0.84	0.80	U	9.14	0.76	0.73	U
AD3	Versatility for use on different types of work sites	0.55	−0.56	U	0.73	0.72	U	8.51	0.62	0.62	U
AD4	Procurement of customised system elements	0.54	−0.55	U	0.71	0.74	U	7.17	0.51	0.53	U

Figure 11. Requirements according to increase in satisfaction/reduction in dissatisfaction.

Figure 11. Requirements according to increase in satisfaction/reduction in dissatisfaction.

4.6. Results of Hierarchisation of Materials According to the List of Requirements

The first step is to form the panel of experts, following the indications established in the methodology. The panel is made up of six experts with no less than 15 years of relevant experience, representing development and a construction companies, engineering and architecture studios, a collaborating mutual society and the body of labour inspectors. Although the sample size is small, several studies have pointed out that panel size is not a limitation, as AHP can be conducted with a small number of participants to achieve sound and statistically robust results [84,85,86]. These experts act as decision makers, making their consensual value judgments on the pairwise comparisons of the modelled AHP problem.

Figure 12 shows the AHP problem presented to the expert panel once the plastic materials and indifferent requirements are discarded.

The second step is to compare the criteria in pairs.

Table A3. Criteria comparison matrix (dimensions).

Criteria	SE	ER	SO	FA	EF	FL	AD	Eigenvector	Weight (Wt)
SE	1	2	5	8	6	9	3	3.868	0.363
ER	1/2	1	4	7	5	8	2	2.726	0.256
SO	1/5	1/4	1	4	2	5	1/3	0.944	0.089
FA	1/8	1/7	1/4	1	1/3	2	1/6	0.337	0.032
EF	1/6	1/5	1/2	3	1	4	1/4	0.652	0.061
FL	1/9	1/8	1/5	1/2	1/4	1	1/7	0.243	0.023
AD	1/3	1/2	3	6	4	7	1	1.883	0.177

Scale: Saaty 1–9; Order: SE ≥ ER ≥ AD > SO ≥ EF > FA ≥ FL; CR = 0.0397 < 0.1.

of Appendix C.1 shows the results (by consensus) of this comparison, as well as the chosen scale and the consistency of the weighting. Next, in

Table A4. Security dimension sub-criteria comparison matrix.

Sub-Criteria	SE1	SE2	SE4	SE5	Eigenvector	Local Wt	Global Wt
SE1	1	7	5	3	3.201	0.564	0.208
SE2	1/7	1	1/3	1/5	0.312	0.055	0.020
SE4	1/5	3	1	1/3	0.669	0.118	0.043
SE5	1/3	5	3	1	1.495	0.263	0.096

Scale: Saaty; Order: SE1 > SE5 > SE4 > SE2; CR = 0.0442 < 0.08.

,

Table A5. Ergonomics dimension sub-criteria comparison matrix.

Sub-Criteria	ER1	ER2	ER4	Eigenvector	Local Wt	Global Wt
ER1	1	1/8	1/5	0.292	0.067	0.017
ER2	8	1	3	2.884	0.661	0.169
ER4	5	1/3	1	1.186	0.272	0.070

Scale: Saaty; Order ER2 > ER4 >> ER1; CR = 0.0420 < 0.05.

,

Table A6. Sustainability dimension sub-criteria comparison matrix.

Sub-Criteria	SU1	SU3	SU5	Eigenvector	Local Wt	Global Wt
SU1	1	1/3	3	0.292	0.067	0.017
SU3	3	1	7	2.884	0.661	0.169
SU5	1/3	1/7	1	1.186	0.272	0.070

Scale: Saaty; Order SU3 > SU1 > SU5; CR = 0.0067 < 0.05.

,

Table A7. Fabricability dimension sub-criteria comparison matrix.

Sub-Criteria	FA1	FA3	FA4	FA6	Eigenvector	Local Wt	Global Wt
FA1	1	3	7	5	3.201	0.564	0.018
FA3	1/3	1	5	3	1.495	0.263	0.008
FA4	1/7	1/5	1	1/3	0.312	0.055	0.002
FA6	1/5	1/3	3	1	0.669	0.118	0.004

Scale: Saaty; Order: FA1 > FA3 > FA6 > FA4; CR = 0.0442 < 0.08.

,

Table A8. Efficiency dimension sub-criteria comparison matrix.

Sub-Criteria	EF1	EF2	EF3	EF4	EF6	Eigenvector	Local Wt	Global Wt
EF1	1	1/3	1/5	3	1/7	0.491	0.063	0.004
EF2	3	1	1/3	7	1/5	1.070	0.138	0.008
EF3	5	3	1	7	1/3	2.036	0.262	0.016
EF4	1/3	1/7	1/7	1	1/9	0.238	0.031	0.002
EF6	7	5	3	9	1	3.936	0.507	0.031

Scale: Saaty; Order: EF6 > EF3 > EF2 > EF1 > EF4; CR = 0.0614 < 0.1.

,

Table A9. Flexibility dimension sub-criteria comparison matrix.

Sub-Criteria	FL1	Eigenvector	Local Wt	Global Wt
FL6	1	1	1	0.023

Scale: Saaty; Order: FL6; CR = 0.0000 < 0.05.

and

Table A10. Adaptability dimension sub-criteria comparison matrix.

Sub-Criteria	AD1	AD2	AD3	AD4	Eigenvector	Local Wt	Global Wt
AD1	1	5	3	9	3.409	0.565	0.100
AD2	1/5	1	1/3	5	0.760	0.126	0.022
AD3	1/3	3	1	7	1.627	0.270	0.048
AD4	1/9	1/5	1/7	1	0.237	0.039	0.007

Scale: Saaty; Order: AD1 > AD3 > AD2 >> AD4; CR = 0.0644 < 0.08.

of Appendix C.2, the sub-criteria (attributes that make up the set of potential requirements) are compared, returning to decide the scale of the comparison and checking the consistency of each matrix.

Table 18. Weighting of criteria and sub-criteria using the AHP method.

ID	Criteria	ID	Sub-Criteria	Sub-Criteria Weight	Criteria Weight
SE	Security	SE1	Resistance to mechanical stress	20.46%	36.29%
		SE2	Resistance to extreme temperatures	2.00%
		SE4	Resistance to bad weather	4.28%
		SE5	Safety through locking parts	9.55%
ER	Ergonomics	ER1	Decomposition into low-weight elements	1.72%	25.58%
		ER2	Assembly/disassembly with ease	16.91%
		ER4	Functionality in different lengths and heights	6.95%
SU	Sustainability	SU1	Use of fire-resistant materials	2.15%	8.85%
		SU3	Use of durable materials	5.92%
		SU5	Availability of spare parts	0.78%
FA	Fabricability	FA1	Manufacturing with light materials	1.78%	3.15%
		FA3	Manufacturing with composite materials	0.83%
		FA4	Manufacturing with recycled materials	0.17%
		FA6	Manufacturing with industrial/standardised materials	0.37%
EF	Efficiency	EF1	Transport by truck	0.39%	6.12%
		EF2	Transport by van	0.84%
		EF3	Transport by car	1.60%
		EF4	Storage in racks	0.19%
		EF6	Procurement at low cost	3.10%
FL	Flexibility	FL6	Use of existing elements in the market	2.28%	2.28%
AD	Adaptability	AD1	Manoeuvrability in different workplaces	9.98%	17.73%
		AD2	Compatibility with existing systems	2.28%
		AD3	Versatility for use on different types of work sites	4.77%
		AD4	Procurement of customised system elements	0.70%

summarises the weights of the criteria and sub-criteria weighted by the panel of experts.

Finally, the panel of experts weighs the adequacy of each alternative for each potential requirement, as reflected in

Table A11. Comparison of alternatives matrix for sub-criterion SE1.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	280/275	280/410	280/465	280/540	1.357	0.262	0.054
AL	275/280	1	275/410	275/465	275/540	1.382	0.267	0.055
CI	410/280	410/275	1	410/465	410/540	0.927	0.179	0.037
GFRP	465/280	465/275	465/410	1	465/540	0.770	0.149	0.030
CFRP	540/280	540/275	540/410	540/465	1	0.747	0.144	0.030

Scale: Nature (yield strength); Order: CFRP (540 N/mm²) > GFRP (465 N/mm²) > CI (410 N/mm²) > GS (280 N/mm²) > AL (275 N/mm²); CR = 0.0018 < 0.1.

,

Table A12. Comparison of alternatives matrix for sub-criterion SE2.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	1/5	1/7	0.491	0.064	0.001
AL	3	1	5	1/3	1/5	1.000	0.130	0.003
CI	1/3	1/5	1	1/7	1/9	0.254	0.033	0.001
GFRP	5	3	7	1	0.33	2.036	0.264	0.005
CFRP	7	5	9	3	1	3.936	0.510	0.010

Scale: Saaty; Order: CFRP > GFRP > AL > GS > CI; CR = 0.0510 < 0.1.

,

Table A13. Comparison of alternatives matrix for sub-criterion SE4.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/5	1/3	1/7	1/9	0.254	0.033	0.001
AL	5	1	3	1/3	1/5	1.000	0.130	0.006
CI	3	1/3	1	1/5	1/7	0.491	0.064	0.003
GFRP	7	3	5	1	1/3	2.036	0.264	0.011
CFRP	9	5	7	3	1	3.936	0.510	0.022

Scale: Saaty; Order: CFRP > GFRP > AL > CI > GS; CR = 0.0510 < 0.1.

,

Table A14. Comparison of alternatives matrix for sub-criterion SE5.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	7	3	2	1.695	0.235	0.022
AL	3	1	9	5	4	3.519	0.488	0.047
CI	1/7	1/9	1	1/5	1/6	0.221	0.031	0.003
GFRP	1/3	1/5	5	1	1/2	0.699	0.097	0.009
CFRP	1/2	1/4	6	2	1	1.084	0.150	0.014

Scale: Saaty; Order: AL > GS > CFRP ≥ GFRP > CI; CR = 0.0473 < 0.1.

,

Table A15. Comparison of alternatives matrix for sub-criterion ER1.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	4.993/12.556	12.073/12.556	4.537/12.556	3.059/12.556	0.507	0.088	0.002
AL	12.556/4.993	1	12.073/4.993	4.537/4.993	3.059/4.993	1.276	0.220	0.004
CI	12.556/12.073	4.993/12.073	1	4.537/12.073	3.059/12.073	0.528	0.091	0.002
GFRP	12.556/4.537	4.993/4.537	12.073/4.537	1	3.059/4.537	1.404	0.242	0.004
CFRP	12.556/3.059	4.993/3.059	12.073/3.059	4.537/3.059	1	2.083	0.359	0.006

Scale: Nature (system weight); Order: CFRP (3.059 kg) ≥ GFRP (4.537 kg) > AL (4.993 kg) > CI (12.073 kg) > GS (12.556 kg); CR = 0.0000 < 0.1.

,

Table A16. Comparison of alternatives matrix for sub-criterion ER2.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	1/5	1/7	0.491	0.064	0.011
AL	3	1	5	1/3	1/5	1.000	0.130	0.022
CI	1/3	1/5	1	1/7	1/9	0.254	0.033	0.006
GFRP	5	3	7	1	1/3	2.036	0.264	0.045
CFRP	7	5	9	3	1	3.936	0.510	0.086

Scale: Saaty; Order: CFRP > GFRP > AL > GS > CI; CR = 0.0510 < 0.1.

,

Table A17. Comparison of alternatives matrix for sub-criterion ER4.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	1/7	1/5	0.491	0.064	0.004
AL	3	1	5	1/5	1/3	1.000	0.130	0.009
CI	1/3	1/5	1	1/9	1/7	0.254	0.033	0.002
GFRP	7	5	9	1	3	3.936	0.510	0.035
CFRP	5	3	7	1/3	1	2.036	0.264	0.018

Scale: Saaty; Order: GFRP > CFRP > AL > GS > CI; CR = 0.0554 < 0.1.

,

Table A18. Comparison of alternatives matrix for sub-criterion SU1.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/5	1/3	1/7	1/9	0.254	0.033	0.001
AL	5	1	3	1/3	1/5	1.000	0.130	0.003
CI	3	1/3	1	1/5	1/7	0.491	0.064	0.001
GFRP	7	3	5	1	1/3	2.036	0.264	0.006
CFRP	9	5	7	3	1	3.936	0.510	0.011

Scale: Saaty; Order: CFRP > GFRP > AL > CI > GS; CR = 0.0510 < 0.1.

,

Table A19. Comparison of alternatives matrix for sub-criterion SU3.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/9	1/7	1/3	1/5	0.254	0.033	0.002
AL	9	1	3	7	5	3.936	0.510	0.030
CI	7	1/3	1	5	3	2.036	0.264	0.016
GFRP	3	1/7	1/5	1	1/3	0.491	0.064	0.004
CFRP	5	1/5	1/3	3	1	1.000	0.130	0.008

Scale: Saaty; Order: AL > GS > CFRP > GFRP > CI; CR = 0.0547 < 0.1.

,

Table A20. Comparison of alternatives matrix for sub-criterion SU5.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	3	5	7	9	3.936	0.510	0.004
AL	1/3	1	3	5	7	2.036	0.264	0.002
CI	1/5	1/3	1	3	5	1.000	0.130	0.001
GFRP	1/7	1/5	1/3	1	3	0.491	0.064	0.000
CFRP	1/9	1/7	1/5	173	1	0.254	0.033	0.000

Scale: Saaty; Order: GS > AL > CI > GFRP > CFRP; CR = 0.0498 < 0.1.

,

Table A21. Comparison of alternatives matrix for sub-criterion FA1.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	4.993/12.556	12.073/12.556	4.537/12.556	3.059/12.556	0.507	0.088	0.002
AL	12.556/4.993	1	12.073/4.993	4.537/4.993	3.059/4.993	1.276	0.220	0.004
CI	12.556/12.073	4.993/12.073	1	4.537/12.073	3.059/12.073	0.528	0.091	0.002
GFRP	12.556/4.537	4.993/4.537	12.073/4.537	1	3.059/4.537	1.404	0.242	0.004
CFRP	12.556/3.059	4.993/3.059	12.073/3.059	4.537/3.059	1	2.083	0.359	0.006

Scale: Nature (system weight); Order: CFRP (3.059 kg) > GFRP (4.537 kg) > AL (4.993 kg) > CI (12.073 kg) > GS (12.556 kg); CR = 0.0000 < 0.1.

,

Table A22. Comparison of alternatives matrix for sub-criterion FA3.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	1/7	1/5	0.491	0.064	0.001
AL	3	1	5	1/5	1/3	1.000	0.130	0.001
CI	1/3	1/5	1	1/9	1/7	0.254	0.033	0.000
GFRP	7	5	9	1	3	3.936	0.510	0.004
CFRP	5	3	7	1/3	1	2.036	0.264	0.002

Scale: Saaty; Order: GFRP > CFRP > AL > GS > CI; CR = 0.0554 < 0.1.

,

Table A23. Comparison of alternatives matrix for sub-criterion FA4.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	7	5	2.036	0.264	0.000
AL	3	1	5	9	7	3.936	0.510	0.001
CI	1/3	1/5	1	5	3	1.000	0.130	0.000
GFRP	1/7	1/9	1/5	1	1/3	0.254	0.033	0.000
CFRP	1/5	1/7	1/3	3	1	0.491	0.064	0.000

Scale: Saaty; Order: AL > GS > CI > CFRP > GFRP; CR = 0.0550 < 0.1.

,

Table A24. Comparison of alternatives matrix for sub-criterion FA6.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/3	3	5	7	2.036	0.264	0.001
AL	3	1	5	7	9	3.936	0.510	0.002
CI	1/3	1/5	1	3	5	1.000	0.130	0.000
GFRP	1/5	1/7	1/3	1	3	0.491	0.064	0.000
CFRP	1/7	1/9	1/5	1/3	1	0.254	0.033	0.000

Scale: Saaty; Order: AL > GS > CI > GFRP > CFRP; CR = 0.0498 < 0.1.

,

Table A25. Comparison of alternatives matrix for sub-criterion EF1.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/2	1	1/3	1/3	0.561	0.099	0.000
AL	2	1	2	1/2	1/2	1.000	0.176	0.001
CI	1	1/2	1	1/3	1/3	0.561	0.099	0.000
GFRP	3	2	3	1	1	1.783	0.313	0.001
CFRP	3	2	3	1	1	1.783	0.313	0.001

Scale: Saaty; Order: GFRP ≈ CFEP > AL > GS ≈ HF; CR = 0.0031 < 0.1.

and

Table A26. Comparison of alternatives matrix for sub-criterion EF2.

Alternatives	GS	AL	CI	GFRP	CFRP	Eigenvector	Local Wt	Global Wt
GS	1	1/2	1	1/3	1/3	0.561	0.099	0.001
AL	2	1	2	1/2	1/2	1.000	0.176	0.001
CI	1	1/2	1	1/3	1/3	0.561	0.099	0.001
GFRP	3	2	3	1	1	1.783	0.313	0.003
CFRP	3	2	3	1	1	1.783	0.313	0.003

Scale: Saaty; Order: GFRP ≈ CFEP > AL > GS ≈ HF; CR = 0.0031 < 0.1.

of Appendix C.3. Figure 13 summarises the results obtained, verifying the adequacy of three alternative materials to the galvanised steel normally used for this type of system: aluminium, glass-fibre-reinforced plastic and carbon-fibre-reinforced polymer. Furthermore, cast iron can be discarded due to its inadequacy in satisfying the requirements contemplated.

5. Discussion of Results

Table 19. Compliance with the tests according to UNE-EN 13374:2013+A1:2019 (model dimensions).

Test	Metals			Plastics		Composites
	GS	AL	CI	PET	PVC	GFRP	CFRP
SLS (deflection)	212%	74%	167%	12%	7%	75%	239%
ULS (bending)	197%	204%	289%	39%	42%	328%	381%
ULS (shear)	13,543%	14,027%	19,831%	2660%	2902%	22,491%	26,119%
PL ULS (bending)	204%	212%	299%	40%	44%	339%	394%
PL ULS (shear)	11,696%	12,114%	17,127%	2297%	2506%	19,424%	22,557%
AL ULS (deflection)	435%	154%	353%	25%	15%	156%	500%
AL ULS (bending)	100%	123%	167%	42%	31%	488%	1164%
AL ULS (shear)	5363%	5363%	7853%	1053%	575%	3831%	4980%

summarises the results obtained in the tests according to the UNE-EN 13374:2013+A1:2019 standard for the selected materials (metals, plastics and composites) and original dimensions (in yellow if the non-compliance is less than 30%, in orange if the non-compliance is higher). Based on these results, plastic materials are discarded due to their low performance in bending tests. On the contrary, previous studies analysed the high-density polyethylene (HDPE) materials to be applied but in surface (non-linear) elements [62]. Likewise, both the reference material and cast iron and carbon-fibre-reinforced polymer are postulated as good candidates for the next stage of the study. However, aluminium and glass-fibre-reinforced polyester require a more detailed study in terms of sizing (outer diameter and/or thickness). If the elements are resized to obtain compliance with the tests, as shown in

Table 20. Sizing of materials to optimise compliance with the tests.

Element	Metals			Plastics		Composites
	GS	AL	CI	PET	PVC	GFRP	CFRP
Guardrail	35−2	35−3	35−2	40−7	40−7	35−4	35−2
Post	40−2	40−2	40−2	40−2	40−7	40−2	40−2

(outer diameter thickness in mm), then the degree of compliance reveals that only plastics have to be discarded, with aluminium and glass-fibre-reinforced plastic becoming suitable, as shown in

Table 21. Compliance with the tests according to UNE-EN 13374:2013+A1:2019 (final dimensions).

Test		Metals			Plastics		Composites
		GS	AL	CI	PET	PVC	GFRP	CFRP
SLS (deflection)	≤55 mm	26	55	33	149	255	55	23
		212%	100%	167%	37%	22%	100%	239%
ULS (bending)	≥281 N·m	555	1019	813	387	422	1634	1071
		197%	362%	289%	138%	150%	581%	381%
ULS (shear)	≥225 N	30,472	59,295	44,619	20,949	22,854	95,076	58,767
		13,543%	26,353%	19,831%	9311%	10,157%	42,256%	26,119%
PL ULS (bending)	≥360 N·m	736	762	1077	387	422	1222	1419
		204%	212%	299%	107%	117%	339%	394%
PL ULS (shear)	≥300 N	35,089	36,342	51,380	20,949	22,854	58,272	67,671
		11,696%	12,114%	17,127%	6983%	7618%	19,424%	22,557%
AL ULS (deflection)	≤300 mm	69	116	85	329	564	115	60
		435%	259%	353%	91%	53%	261%	500%
AL ULS (bending)	≥781 N·m	7ype85	1701	1308	1161	851	6764	9093
		100%	218%	167%	149%	109%	866%	1164%
AL ULS (shear)	≥625 N	33,519	62,975	49,081	23,044	12,570	44,982	31,125
		5363%	10,076%	7853%	3687%	2011%	7197%	4980%

. The use of aluminium alloys in temporary demountable structures has been previously analysed [87]. However, to the best of our knowledge, the use of composite materials such as GFRP or CFRP is a novelty in the field of collective safety protection on construction sites.

As for the requirements to be considered, the Kano method and its variants allow the requirements from the literature review to be classified, with the indifferent ones being discarded.

Table 22. List of attractive, mandatory and one-dimensional requirements.

ID	Requirement	Type
SE2	Resistance to extreme temperatures	Mandatory
SE4	Resistance to bad weather
SU1	Use of fire-resistant materials
FA6	Manufacturing with industrial/standardised materials
EF1	Transport by truck
EF6	Procurement at low cost
SE1	Resistance to mechanical stress	One−dimensional
ER1	Decomposition into low-weight elements
ER2	Assembly/disassembly with ease
ER4	Functionality in different lengths and heights
SU3	Use of durable materials
EF2	Transport by van
EF4	Storage in racks
FL6	Use of existing elements in the market
AD1	Manoeuvrability in different workplaces
AD2	Compatibility with existing systems
AD3	Versatility for use on different types of work sites
AD4	Procurement of customised system elements
SE5	Safety through locking parts	Attractive
SU5	Availability of spare parts
FA1	Manufacturing with light materials
FA3	Manufacturing with composite materials
FA4	Manufacturing with recycled materials
EF3	Transport by car

shows the mandatory, attractive and one-dimensional requirements according to the results obtained through a survey of 190 potential users. In this way, of the 42 initial requirements, 24 stand out from the rest. It should be noted that all defined dimensions (safety, ergonomics, sustainability, fabricability, efficiency, flexibility and adaptability) are represented.

The capacities of the system to withstand extreme temperatures (including fire) and inclement weather, to be manufactured in an industrialised way so that it can be transported by lorry and to be of low cost are postulated as mandatory requirements capable of reducing dissatisfaction. Likewise, for the system to have high mechanical capacity and durability, to be made up of lightweight elements, to be easy to assemble and disassemble in different lengths and heights and to be transportable by van and storable on shelves, as well as the possibility of acquiring these elements independently, are postulated as one-dimensional requirements capable of both reducing dissatisfaction and increasing user satisfaction. In addition, the system being able to use elements that already exist on the market and being compatible with other current systems, as well as its manoeuvrability in different workplaces and its versatility for different types of work, are also postulated as one-dimensional requirements. Finally, the attractive requirements that stand out are the incorporation of safety measures such as using locking parts, the availability of spare parts, the use of lightweight composite, recycled and/or recyclable manufacturing materials, the ability to be transported in a private passenger car, and all are capable of increasing the satisfaction of potential users. These findings are aligned with previous studies on scaffolding standardisation [88].

Once the requirements are classified, they are hierarchised using the AHP method without differentiating their ability to increase satisfaction and/or reduce user dissatisfaction on incorporation, establishing a pairwise comparison system at two levels: criteria and sub-criteria. This process is summarised in Figure 14.

With the first eight requirements, 79% of the decision is made (mechanical strength, ease of assembly/disassembly, manoeuvrability in different workplaces, safety through locking parts, functionality in different lengths and heights, use of durable materials, versatility to be used in different types of work and resistance against inclement weather). Furthermore, with the following eight requirements, 95% of the decision is reached (low-cost acquisition, incorporation of existing elements on the market, compatibility with current systems, use of fire-resistant materials, resistance to extreme temperatures, manufacture with lightweight materials, breakdown into lightweight elements and transport by private car). The remaining eight requirements complete the decision (transport by van, manufacturing with composite materials, availability of spare parts, personalised acquisition of system elements, transport by lorry, industrial/standardised manufacturing, shelf storage and manufacturing with recycled and/or recyclable materials). The assessment, in weighted terms, of each alternative to align with each attribute is summarised in Figure 15, where the reference material and the three most outstanding alternative materials are shown.

Whereas the purpose of the Kano model is to distinguish the needs of users through different processes, helping to find the point for improving satisfaction and mitigating dissatisfaction, the AHP method stresses the most relevant basic, functional and excitement needs that have been previously underlined. In summary, the Kano model compiles needs in three basic, functional and motivational categories; the AHP method provides a hierarchical classification, highlighting the most appropriate alternatives. This integration is aligned with previous studies that combined both methodologies [36,37,38,39].

6. Conclusions

Advances in the legislative field related to occupational health and safety in the construction sector have not caused a proportional decrease in occupational accidents. This has been due, in part, to the failure to bring about the appropriate technical changes to allow for the correct application of the standards. This can be explained by the increase in the amount of building work over the years of the real-estate bubble in Spain and the decrease in investment during the years of subsequent crisis but also by a stagnation in the development of new and better systems in terms of collective protection, especially in the field of falls from height, so that these systems cease to be defective and/or misused and their use is encouraged.

A series of needs to be satisfied and requirements to be met was provided, duly compiled, prioritised and hierarchised. At this point of the study, four alternative materials to galvanised steel were proposed to be used to lay the foundations for an innovative preliminary design of TEPS. From the materials with potential as an alternative material for new TEPS, aluminium alloy 2024 and cast iron (as metals) and glass-fibre-reinforced polyester and carbon-fibre-reinforced polymer (as composite materials) were selected. In addition, the Kano model enabled requirements (those considered as one dimensional and mandatory) and needs (those considered as one dimensional and attractive) to be filtered. Through the AHP multi-criteria decision-making process to be developed from this study, alternative materials can be hierarchised in relation to the criteria that the Kano model has provided.

It can be noted that the proposed approach faces each requirement individually, so it does not consider the different interrelations among customer needs. In addition, although this research provides several inputs to design and develop innovative solutions, future research faces several challenges and constraints. Therefore, future research must involve the preliminary design of TEPS, taking into account their geometric and dimensional definition in order to study their patentability and/or usefulness. After that, the detailed design, development and prototyping of TEPS can be performed. Finally, the validation of solutions for their potential commercialisation should be addressed by conducting market research surveys of users.