What Drives Acceptance of Occupational Exoskeletons? Focus Group Insights from Workers in Food Retail and Corporate Logistics

Abstract

The potential of occupational exoskeletons can only be realized if workers are willing to wear them on their bodies. As classical technology acceptance theories originate in research into information technology, they do not sufficiently cover the peculiarities of exoskeletons, and thus greater focus is needed on factors that specifically shape intentions to use them. Involving three companies from food retail and logistics, we conducted guided focus groups with 18 workers who perform material handling tasks in their daily work. Participants discussed the envisioned benefits, risks, and conditions related to the adoption of exoskeletons. Consistent with established technology acceptance models, performance-related and effort-related factors were found to be highly recognized. Participants also highlighted factors such as wellbeing, fairness, and the altered physical appearance of wearers, which was considered important in social contexts at work. Complementing established factors with new exoskeleton-specific determinants, our results are a valuable starting point for further exoskeleton user studies.

1. Introduction

Despite advances in industrial automation, many jobs continue to require human workers to perform physically demanding tasks (Eurofound, 2019; Mital & Pennathur, 2004). Repetitive movements, forced body postures, and lifting and carrying of heavy loads pose significant strain on the musculoskeletal system and are therefore considered health risk factors (Da Costa & Vieira, 2010; de Kok et al., 2019; Norman et al., 1998). In recent years, new assistive wearable devices have been developed to help workers in executing material handling tasks while maintaining their physical mobility (de Looze et al., 2016). Research has provided evidence that the use of industrial exoskeletons – as this new type of body-worn technology is termed – can reduce biomechanical stress or strain, for example, in logistics workers who often need to bend forward or lift asymmetrically (Bär et al., 2021). Depending on their activation principles, exoskeletons generate assistive forces differently and can therefore be classified as active, passive or a combination thereof (Lowe et al., 2019; Toxiri et al., 2019). Passive systems use mechanical mechanisms, such as springs, levers, dampers, and special fabric to redistribute kinetic energy. In contrast, active devices employ powered actuators, such as electric motors, to augment a wearer’s naturally given strength beyond human capacity. Robotic exoskeletons can also integrate various sensors or microcomputers that are supposed to interact with the user. Algorithms, for example for EMG-based pattern recognition and control strategies, might be applied to acquire information as a prerequisite for the execution of a motor function (Ruiz-Olaya et al., 2018). In view of these developments, exoskeletons are an increasingly relevant subject for HCI research.

Four systematic reviews of active and passive exoskeletons for back support and upper limb assistance have compiled data that indicates, for example, a potential decrease in muscle activity in body regions supported by an exoskeleton (de Looze et al., 2016; Kermavnar et al., 2021; McFarland & Fischer, 2019; Theurel & Desbrosses, 2019). These studies suggest that use of exoskeletons during occupational tasks at least modifies physical strain and therefore has the potential to counteract the risk of developing musculoskeletal complaints (Steinhilber, Luger, et al., 2020). In order to unfold the health-promoting potential of exoskeletons, however, workers must be willing to consistently wear them during task execution. Given the relative novelty of exoskeletons in occupational settings and their unique requirement to be worn on the body, research into the conditions of user acceptance is urgent. Understanding target users’ thoughts and attitudes, perceived benefits, and risks related to occupational exoskeletons allows a more holistic picture to be drawn of what can foster or reduce people’s willingness to start and continue using exoskeletons at work.

Exploring perceptions of potential exoskeleton users – including not only performance-related and utilitarian, but also social and emotional aspects – may help to critically reflect on possible effects on the wellbeing of workers when the technology is to be implemented in organizations (Glock et al., 2020). Since the design of industrial exoskeletons can vary considerably depending on the materials and mechanisms used to construct them, end users’ perceptions can also be an expedient route to a better understanding of critical design requirements. Reflecting opinions and user perspectives can even provide guidance for the evaluation of a user’s real-life experience (Kim et al., 2022; Pesenti et al., 2021). Former studies which included subjective evaluations of exoskeletons found that perceived musculoskeletal discomfort or poor wearing comfort can diminish exoskeleton acceptance or even lead to the rejection of a device (Dewi & Komatsuzaki, 2018; Hensel & Keil, 2019; Siedl et al., 2021). However, reasons why people choose or decline to wear an exoskeleton can go beyond feeling comfortable, as qualitative findings from the field of medical rehabilitation (Baltrusch et al., 2020) and research on older adults (Shore et al., 2020) illustrate. For patients suffering from low-back pain, for example, individual and gradual adjustability, independence in donning and doffing, and evidence of long-term positive efficacy of exoskeletons turned out to be key requirements when considering use of wearable lifting devices (Baltrusch et al., 2020). Older adults additionally found the noise generated by exoskeletons to be relevant and whether the devices enable users to carry out control and basic maintenance autonomously (Shore et al., 2020).

Placing an explicit focus on the work context, psychological research into technology acceptance offers a mature starting point for shedding light on factors that are crucial to predicting behavioral intentions towards using a technology (Hornbæk & Hertzum, 2017). However, since theories and research models dealing with technology acceptance, such as the Technology Acceptance Model (TAM) 1–3 (Davis, 1989; Venkatesh & Bala, 2008; Venkatesh & Davis, 2000) and the Unified Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh et al., 2003), originated in the field of information technology, it remains unclear to what extent they take account of the specifics of exoskeleton technology and the interaction between user, device, and organizational environment. The aim of this study was therefore to identify factors that influence workers’ intention to use industrial exoskeletons and to detect deviations from relevant theoretical models. Scheel et al. (2021) suggested that gaining profound field knowledge is crucial for theory and hypothesis building. To this end, qualitative research methods such as focus groups bring the researcher closer to potential users and help to figure out their beliefs, needs, and opinions (Veling & McGinn, 2021). Since task specificity was generally found to play an important role in exoskeleton assessment (Alemi et al., 2020), we involved participants from various organizations in focus group discussions, covering wide ranges of logistical tasks and organization-specific conditions of task execution.

2. Method

Focus groups as a qualitative research method allow participants to present, explain, and discuss their individual views in an interactive setting to identify common and opposing opinions, and to discover underlying perceptions through collaborative discourse (Millward, 2012). We conducted three focus groups (N = 5; 6; 7) to understand what cognitions, assumed risks, and benefits come to workers’ minds when being confronted with exoskeletons as future assistive devices at work. In doing so, we aimed to identify indications of which influencing variables - potentially going beyond those in established technology acceptance models - should be considered in research into the acceptance of exoskeletons.

2.1. Participants

We recruited workers from three Austrian industrial companies, two operating in food retail and one in fire truck manufacture, where deployment of exoskeletons was being considered. Although all participants were engaged in physical material handling tasks as part of their work activity, an effort was made to include individuals from as many different workplaces as possible. The first focus group consisted of four women and one man whose work tasks in the supermarket involved receiving goods from suppliers, unpackaging goods and stocking shelves, cashier work and customer service. The second focus group consisted of six participants (three male, three female), four of whom reported a focus of daily work tasks on shelf stocking, one worked in the pastry department and one was part of the cleaning staff. Seven male logistics employees of a fire truck manufacturer formed the third focus group, with work responsibilities for order picking and inventory work, and storage of incoming goods and their transportation to other departments. Across all groups (N = 18), the workers were aged 19 to 53 (M = 31.72 years, SD = 10.37). 39% classified themselves as female, 61% as male, and 0% as non-binary. The majority of participants reported having little or no prior experience with exoskeletons in practice (M = 1.44, SD = 0.70). For an overview of participant characteristics see Table 1.

Table 1. Characteristics of participants.

Focus group	Participant code	Gender	Age	Work area	Work tasks
1	BI_M1	Female	53	Deputy store manager	RG, SS, CW, and CS
1	BI_M2	Female	38	General Duties	SS, CW, and CS
1	BI_M3	Female	22	Deputy store manager	RG, SS, CW, and CS
1	BI_M4	Female	47	Store manager	RG, SS, CW, and CS
1	BI_M5	Male	20	Apprentice	SS, CW, and CS
2	ME_M1	Female	49	Cleaning service	Cleaning work
2	ME_M2	Male	25	Non-food goods department	SWW
2	ME_M3	Male	29	Refrigeration department	SWW
2	ME_M4	Male	39	Refrigeration department	SWW
2	ME_M5	Female	19	Pastry shop	Pastry preparation
2	ME_M6	Female	20	Refrigeration department	SWW
3	RB_M1	Male	28	Logistics department	Internal transport
3	RB_M2	Male	23	Logistics department	OPIW
3	RB_M3	Male	37	Logistics department	OPIW
3	RB_M4	Male	29	Logistics department	OPS in high rack storage
3	RB_M5	Male	32	Logistics department	OPS on ground level
3	RB_M6	Male	35	Logistics department	OPS in middle section storage
3	RB_M7	Male	26	Logistics department	OPIW

Note. RG: receiving goods from suppliers; SS: stocking shelves; CW: cashier work; CS: customer service; SWW: shelving and warehouse work; OPIW: order picking and inventory work; OPS: order picking and stocking.

By sending out an information letter, we invited all potential participants to take part in the study. Those who were interested signed up voluntarily for participation directly at their companies. Our research complies with the tenets of the Declaration of Helsinki and was approved by the work council committees of all participating companies.

2.2. Procedure

One researcher (S.S.) moderated the focus group discussions in German language, using a previously defined interview guide (see Figure 1). A second person observed the discussion setting, took notes, and documented any additional remarks. Before the focus group discussions started, participants provided demographic information in a short questionnaire and gave written informed consent. Workers introduced themselves in terms of their workplaces and associated tasks. Afterwards, we presented four 3-min videos to participants, showing various exoskeletons in use (Chairless Chair by Noonees, V22 ErgoSkeleton by StrongArm Tech, PLAD by PeakWorks, Paexo Shoulder by ottobock) and briefly explaining the devices. The stimulus material was intended to convey a clear impression of the purpose and functionality of exoskeletons and to illustrate what they can look like. We stressed that the exoskeletons presented were merely examples of systems already on the market and that the questions to be discussed subsequently were aimed to elicit personal views on the acceptance of industrial exoskeletons more generally. With permission of all participants, the focus group discussions were fully audio-recorded. They lasted between 50 and 60 min and ended with an outlook on what would happen with the data gained during the sessions.

Figure 1. Process and key questions of focus group meetings.

2.3. Data analysis

The audio recordings were transcribed verbatim, and German colloquialisms and grammatical errors were subsequently corrected. The overall word count of transcripts amounted to 19,793. For reasons of anonymity, the names of participants and their corresponding organizations were replaced by numbers. Against the background of our research question, we followed the approach of qualitative content analysis (Schreier, 2012) as a method to systematically describe the meaning of the data by its classification into categories of a theory-based coding framework. After the moderating researcher (S.S.) had carefully read through the transcripts, the coding process continued by identifying and assigning text snippets (181 coding units) to various subcategories which specified dimensions of the key focus of analysis. The coding, its subcategories, and preliminary dimensions were subject of several rounds of profound discussion (with M.M and A.P.) until consensus was reached.

3. Results

We intended to capture diverse views and perceptions regarding the idea of using an exoskeleton in the workplace. Since the participants all had in common that they performed various kinds of manual material handling tasks, the data gathered in all three groups was analyzed collectively using the coding framework developed. We built this coding framework against the background of constructs of established models of technology acceptance (Davis, 1989; Goodhue & Thompson, 1995; Venkatesh et al., 2003; Venkatesh & Bala, 2008) while also making space for new categories not previously considered therein. Peculiarities with respect to individual work areas were explicitly highlighted. “Fit of technology”, “performance-related effects”, and “effort” were identified as key dimensions that are very well represented in the above-mentioned models and theories. Relevant to the acceptance and actual use of exoskeletons, but hitherto largely neglected, “health and wellbeing effects”, “appearance and attractiveness”, and “social acceptability” were added as further main dimensions from our focus group data (for an overview see Figure 2).

Figure 2. Structured illustration of focus group results.

3.1. Fit of technology

In order to gauge the applicability of an exoskeleton to individual users in a specific workplace, participants in all groups started with pointing to criteria relating to a more general fit of technology as prerequisites for exoskeleton usage.

3.1.1. Task-technology fit

With reference to their own jobs, participants’ discussions revolved around the support functions attributed to exoskeletons and whether they might fit the respective profiles of their workplaces and workplace conditions. Based on their expertise in current work processes and task accomplishment, they mentioned potential challenges of using specific exoskeletons:

The problem is because the exoskeleton is designed for certain tasks, but we perform different activities; you would need 15 different exoskeletons. […] (RB_M4, Z53–55)

This also included compliance with safety standards.

[…] With straps and belts sticking out, the greatest danger is that you can get caught on the pallets. …] (RB_M6, Z248) […] It is not by chance that we are not allowed to wear bracelets. (RB_M4, Z88)

3.1.2. Person-technology fit

Participants also discussed the necessity of achieving a proper fit between technology and individual body dimensions, and the challenge associated with sharing an exoskeleton between colleagues with great differences in body proportions.

If I compare him [points to colleague] with me, […] we could not wear the same exoskeleton. (RB_M6, Z277–279)

Additionally, the level of support a worker needs can vary between individuals, depending on workplace characteristics and personal physical constitution. Consequently, the perceived personal relevance of using an exoskeleton differs interindividually and seems to be linked to the willingness to use an exoskeleton at work.

If I do not have any physical complaints, I do not think I need it. (ME_M1, Z434–435)

Further, they identified certain groups of employees who would benefit most from exoskeleton use.

The young probably do not want to put it on, but the older ones, like me, they would wear it because they are in pain. (BI_M1, Z325–332)

Overall, participants emphasized the need for personal experience to assess whether an exoskeleton is appropriate for individual wearers and their work conditions. One participant simply summarized this as follows:

[…] You would really have to test the devices directly on site so that you can say how it is, how it works, how it feels. […] (RB_M2, Z73)

3.2. Work-performance-related effects

Participants emphasized that using an exoskeleton must support and not inhibit a worker's job performance. The majority of participants expressed concern that an exoskeleton could interfere with their efficiency and decrease work productivity. Aspects of flexibility and work speed were highlighted.

3.2.1. Flexibility

Since a variety of movements were required in all work areas represented (e.g., bending down/forward, frontal/lateral/asymmetric lifting), flexibility was mostly understood in terms of unrestricted movement while wearing the exoskeleton.

[…] It is important that I can move freely, but with such big, bulky systems? Something like that would just hinder me. […] (ME_M1, Z455–456)

3.2.2. Work speed

With respect to the intense time pressure that retail workers are under in daily business, the requirement of speed in executing tasks while wearing the exoskeleton emerged strongly from the first two focus groups. Retail workers reported having developed efficient work routines that enable them to manage large work volumes. As a result, exoskeletons should allow both these routines and their speed to be maintained.

When you get 2.5 tons of goods delivered, you have to be fast with your hands. […] And that is why we need something that you can move really fast with. (BI_M1, Z173–176)

3.3. Health and wellbeing effects

Talking about the benefits and risks of industrial exoskeletons, participants expressed clear ideas about positive consequences for personal health and wellbeing, but also addressed potential threats and disadvantages.

3.3.1. Relief potential

Referring more generally to work activities that involve high physical workload, participants stated that a non-restrictive correction in body posture and effective physical support in certain body regions, such as the lower back and the shoulder area, would be desirable. Although preservation of long-term health was mentioned as an issue, tangible short-term effects in terms of noticeable reduction in work-induced physical pain and perceptible physical relief received most attention.

I would expect to be pain-free. I come home without feeling pain. That would be great. (BI_M1, Z194)

Participants in the third group, however, questioned a “purely” positive effect with respect to not yet foreseeable consequences in losing muscle strength and diminishing mobility of human joints when using the exoskeleton in the long run.

What happens when you perform a work task with the exoskeleton eight hours a day? I assume that a development is triggered that leads to some kind of atrophy. (RB_M7, Z9–10)

3.3.2. Comfort and hygiene

In all three focus groups, participants stressed the importance of feeling physically comfortable when wearing an exoskeleton. One participant put it as follows:

The exoskeleton should be comfortable. This means it does not scratch, press or pull anywhere. […] If it is not comfortable, then you do not like to wear it. (BI_M5, Z124–125)

Moreover, the additional weight of the exoskeleton, specific clothing requirements (e.g., the need to wear a jacket over the exoskeleton when entering a freezer), and effects of high temperatures such as increased perspiration and heat sensation played a role in discussions about wearing comfort.

[…] We do not have air conditioning in our workplace, and it is extremely hot. I cannot imagine wearing this [the exoskeleton] additionally. In winter, it would be ok. (BI_M1, Z457–458)

Specifically, given higher temperatures and a shared use of devices, participants expressed discomfort regarding hygiene and cleanliness.

Besides, it is not very hygienic when the previous wearer has sweated […]. I would refuse to take the same then. (RB_M6, Z289–292)

3.3.3. Feeling of safety

Discussions also centered on the questions of how well and stably industrial exoskeletons really work and how reliable and robust they appear to be under specific working conditions (dust, differences in temperature, etc.). Although none of the participants expressed personal fear of being hurt by an exoskeleton, the risk of accidents at work was addressed, particularly in relation to the structure and components of exoskeletons and becoming caught in some way.

We have a small freezer; there are ten thousand boxes in there, and when I am at the top of the ladder, I cannot bend down or anything, so I would get stuck on the shelf with the exoskeleton. (ME_M5, Z1682–684)

3.4. Effort

Participants agreed on the fact that additional effort associated with the use of an exoskeleton had to be kept as low as possible. Several potential causes for an increase in effort were mentioned:

• Increased cognitive effort may occur if the exoskeleton requires users to pay extra attention to their immediate work environment and to their own conscious movements.

• Donning and doffing the exoskeleton and individual adjustments may constitute a second source of effort, and should therefore be easy, require minimal time effort, and allow independent device handling (no requirement to be supported by others).

• Usage of an exoskeleton may require its wearers to take on responsibility: to keep it intact, appropriately stored (to prevent device loss) and clean from dirt, dust, and sweat. Thus, several participants considered the availability of an adequate organizational infrastructure necessary.

With reference to adjustment and responsibility issues, participants expressed skepticism particularly about the sharing of systems between workers.

[…] If it takes you ten minutes to correctly adjust the system, that is time you lose working. It would be better if everyone used their own system. (RB_M3, 286–288)

3.5. Appearance, attractiveness, and social acceptability

While the third group did not bring up visibility and aspects of aesthetics, retail workers approached these topics actively and comprehensively within their groups. Participants discussed what other people, for instance, customers entering the shop, would think when they saw workers wearing exoskeletons, and how this would influence their behaviors and reactions. Furthermore, participants brought up a discussion of fair use of exoskeletons in the workplace.

3.5.1. Visibility

As retail workers collectively seek to avoid attracting attention, they expressed a clear preference for an inconspicuous, streamlined design without bulky elements that could be worn ideally over regular clothing but covered with a jacket or other kind of workwear. One participant explained that being seen with the device by customers would be an impediment because

[T]hey will ask what we are wearing and why we are doing it. They have 100 questions. (BI_M2, Z425)

Another participant remarked that

[…] We also have many children in the store and they will stop us. They will just want to play with us. […] (ME_M3, Z149–151)

3.5.2. Aesthetics

Participants also talked about personal appearance and argued that visible devices would make them look funny, weird or less human, with a negative impact on perceived personal attractiveness. One participant expressed his concern that, as a deputy store manager, he would appear less trustworthy and lose credibility with customers.

[…] I think it would look a bit untrustworthy if I talked to customers in my shirt while wearing such a device. […] (ME_M2, Z194–195)

He continued, with a colleague chiming in:

It does not really seem human to me […] (ME_M2, Z198)

It looks like a robot. (ME_M4, Z198)

3.5.3. Subjective norm and social context

Perceptions by customers, colleagues, and supervisors were regarded to be relevant insofar as it seemed to matter to participants what these groups of people would think of them when using an exoskeleton. To see employees using exoskeletons might, on the one hand, give customers an impression of how stressful work in food retail actually is and thus lead to more external understanding and recognition and a more positive company image.

They would think that [the company] is making our hard work easier. I believe that this will be taken positively by customers. (BI_M4, Z439–440)

On the other hand, exoskeleton wearers might experience some kind of devaluation if perceived less as sentient human beings and as intimidating by others. Two other participants remarked that wearing an exoskeleton could also be interpreted as an indication of injury or health problems and trigger stigmatization or disrespect by colleagues and others.

[…] If a person looks like that, they will be made fun of [by their colleagues]. (ME_M3, Z658–659)

In order to avoid judgement by the social environment and escape social reactions, one retail worker mentioned that

[…] I can imagine to wear the exoskeleton, but only when I am receiving goods alone in the storage area. (BI_M1, Z105–106)

3.5.4. Fairness

Discussing the case that not every employee is given their own exoskeleton, participants focused on conditions and consequences of a fair distribution of exoskeletons among workers. A variety of views were expressed regarding criteria for the personal availability of an exoskeleton.

[…] If you are really in pain, you should be allowed to wear the exoskeleton without anyone saying no. (BI_M1, Z338–339)

To those who execute the most strenuous work I would give the exoskeleton. […] (BI_M4, Z349)

Participants agreed that a company-wide distribution of exoskeletons – whether perceived as fair or unfair, a privilege or a disadvantage – would benefit social dynamics. A single person would no longer stand out. This would minimize grounds for social exclusion of individual exoskeleton wearers.

4. Discussion

Our focus group study is a further step towards identifying criteria that drive the adoption of exoskeletons in work environments. We explored in-depth the perceptions of workers who reflected on whether they would want to use industrial exoskeletons themselves. Their statements were analyzed with reference to the extensive literature on technology acceptance, specifically TAM1–3 (Davis, 1989; Venkatesh & Bala, 2008; Venkatesh & Davis, 2000), UTAUT (Venkatesh et al., 2003) and the Task-Technology-Fit Model (TTFM) (Goodhue & Thompson, 1995). Therein established determinants, that is, expected effects on performance and effort, were also well reflected in our qualitative data. Beyond such utilitarian factors, however, our research contributes to a more holistic understanding of exoskeleton acceptance by pointing to additional criteria, such as appearance, attractiveness, and social acceptability. Although these factors have hitherto received little to no attention in the research on and the design and application of exoskeletons, they probably do not only impact behavioral intentions to use exoskeletons, but also workers’ wellbeing. Consequently, these aspects should be considered when implementing exoskeletons in organizations. Below, we discuss our findings with reference to the published literature and practical implications.

As participants emphasized, applicability of exoskeletons in everyday professional life will require both task-technology fit (i.e., exoskeleton suitability for specific workplaces and tasks) and person-technology fit (i.e., exoskeleton suitability for individual needs or body proportions). This is in line with recommendations regarding the general use of exoskeletons in the occupational context developed by Steinhilber, Luger, et al. (2020). Previous studies that investigated reductions in mechanical load on the musculoskeletal system illustrated that health effects of exoskeletons differ between work tasks and individual ways of task execution (Alemi et al., 2020; Madinei et al., 2020; Steinhilber, Bär, et al., 2020). Further, according to the TTFM by Goodhue and Thompson (1995), a good interplay between person, technology, and task characteristics is crucial. Investigating wearable devices (smartwatches), Al-Emran (2021) was able to show an influence of individual-technology fit and task-technology fit on perceived usefulness. Further empirical work underlines the relevance of fit for technology, indicating a positive correlation between perceived task-exoskeleton fit and user acceptance (Siedl et al., 2021). In practice, thorough workplace analysis and the involvement of workers with detailed knowledge of their needs and workplaces should therefore precede exoskeleton implementation. Since participants made close connections between personal experience (the need to test devices directly at work) and behavioral intentions to use an exoskeleton, organizational pretesting is recommended.

Participants were concerned about using exoskeletons having negative effects on work performance due to presumed reductions in flexibility and work speed. In line with the definitions of usefulness in TAM1–3 and performance expectancy in UTAUT – both of which clearly focus on a technology’s potential to improve job performance – our results reflect the importance of this utilitarian element. Interestingly, the focus was more on preventing a drop in work performance (e.g., loss in speed) than on achieving an effective increase in job performance (e.g., working faster). Consideration of (design) characteristics that influence these aspects therefore seems to be an option for improving the acceptance of exoskeletons in the workplace. Comparing rigid exoskeletons made of solid and stiff components to soft exoskeletons made of fabric structures, Siedl et al. (2021) found rigid systems more likely to be associated with a slowdown in work speed.

Furthermore, we found that the lack of tangible physical strain relief and ease of pain and low assumed comfort decreased workers' intention to use exoskeletons. Some participants even expressed fears about harmful physical side effects. In contrast to such perceptions, however, exoskeletons, like other assistive technologies for manual material handling, are designed to reduce wearers' work-related stress and thus increase their wellbeing (Glock et al., 2020). One could therefore argue that positive health effects also serve as an important performance indicator that is typically not represented in traditional models of technology acceptance. Besides, feeling physically relieved can strengthen work-related self-efficacy beliefs (Siedl & Mara, 2021), that is, how capable a worker feels of successfully accomplishing a task. Integration of subjective evaluations and psychological measures could thus add valuable insights to future exoskeleton studies.

Our participants also associated exoskeletons with additional effort in terms of increased cognitive load (e.g., having to pay more attention to the work environment), responsibility, and device handling (donning, doffing, and adjusting). In research into an adapted version of UTAUT (Dwivedi et al., 2019), effort expectancy turned out to be highly predictive of intentions to start using exoskeletons among workers (Elprama et al., 2020). However, the operationalization of this construct (also see “ease of use” in TAM1–3) (Venkatesh et al., 2003) focuses on a more general dimension (e.g., how difficult it is to learn to operate the system or how easy it is to handle the device) anchored in the interaction between user and technology, but does not consider mundane organizational aspects that our participants came up with (Shachak et al., 2019). When it comes to exoskeletons as assistive devices in the workplace, established constructs thus fall short, as they are too undifferentiated to capture the various presumed sources of additional effort and to derive interventions for system design and implementation (e.g., infrastructure for storage, maintenance, and cleaning of exoskeletons).

Interestingly, our data also revealed that the visibility of exoskeletons on the wearers’ bodies is expected to provoke social reactions from colleagues or customers which are linked to effects on work performance, effort, and psychosocial wellbeing. Beyond a system’s visibility, the idea of one's own appearance as exoskeleton user triggered participants to think about what they would look like and how others would perceive them. Whereas the design of exoskeletons can alter attributions (e.g., human vs. more machine-like) to technology users (also see Meyer & Asbrock, 2018), the social context seems to influence the judgment of others and determines the consequences that result from a user’s interaction with a technology (Uhde & Hassenzahl, 2021). Therefore, taking adequate measures to avoid negative group dynamics when implementing exoskeletons in organizations is crucial. According to our focus groups, visibility, aesthetics, and social context can be co-decisive for the acceptance or rejection of exoskeletons. This might be particularly true if the work area is open and accessible to others. Since most of these aspects are not yet integrated into traditional models of technology acceptance, simply applying these models to occupational exoskeletons would fail to take into account the specifics of wearable – and visible – technology.

There are parallels between relevant literature on exoskeleton evaluation and feedback from our focus group participants. In a systematic review, Pesenti et al. (2021) elaborated on potential standard criteria for performance validation of spinal exoskeletons, introducing five ways of evaluating the effectiveness in human support, including (1) EMG-based measurement of muscle activity, (2) computation of compression forces on relevant human joints or their flexion-extension moment, (3) quantification of the metabolic cost/rate, (4) assessment of the functional/task-related performance (e.g., measurement of restrictions in human movement or execution time), and (5) capturing the subjective user experience. The fifth domain, which focuses on the wearer's perception, integrates subjective feedback from users on how comfortable, useful, supportive, or strain-relieving an exoskeleton might be. In accordance with some health and wellbeing effects envisioned by our focus group participants, subjectively perceived changes in physical exertion and task difficulty were found to be crucial for the adoption of industrial exoskeletons (Pesenti et al., 2021; Siedl et al., 2021). Immediate and observable benefits, such as a noticeable reduction in physical pain, seem to matter across industries, as do expectations of positive long-term effects and safety concerns, such as unexpected failure and risk of injury (Cha et al., 2020; Upasani et al., 2019). Further concerns are performance-related effects which coincide with functional metrics, such as range of motion/freedom of movement and work speed as an indicator of task success (Pesenti et al., 2021; Upasani et al., 2019). Since the physical support offered by a specific exoskeleton differs by task (Kozinc et al., 2020), previous research has already emphasized device versatility as an essential requirement for exoskeleton acceptance (Baltrusch et al., 2020). A highly versatile exoskeleton interferes little with the work of its user, even in functional tasks such as ladder climbing, where an exoskeleton might restrict user movement (Baltrusch et al., 2018). Hence, task variability – a characteristic of many real-world work settings – must be represented adequately when assessing user acceptance of exoskeletons. The systematic twelve-task approach devised by Kozinc, Baltrusch, et al. (2021) to investigate the effects of a passive trunk exoskeleton on functional performance, discomfort, and user satisfaction, may thus offer a model for tackling the challenge of multi-task end user testing. Focus group results further indicated that testing in the field (i.e., use of an exoskeleton in a specific workplace and/or organizational environment) plays a pivotal role in fully capturing a wearer’s exoskeleton experience.

In contrast to controlled trials in the laboratory, end user testing in the real world has the potential to reveal effects of social perception and context: Visibility and personal appearance turned out to be particularly meaningful to users when envisioning third-party reactions at work (e.g., by customers). This is accords with previous research that acknowledged the relevance of design features (e.g., color, medical vs. industrial style) in exoskeleton acceptance and the influence of people’s culturally driven attitudes and beliefs (Baltrusch et al., 2021; Cha et al., 2020).

According to McMillen and Söderberg (2002), people with disabilities tend to accept assistive devices only when the technology serves their individual purposes. As indicated by our participants, to workers without medical issues, the reason for using an assistive exoskeleton may not necessarily be apparent. Recent studies by Kozinc, Babič, and Šarabon (2021) and Baltrusch et al. (2021) reported that people with low-back pain are more willing to use back-assistive exoskeletons than healthy individuals. A limited willingness to use an exoskeleton in the absence of physical pain and discomfort is consistent with the results of our study, and highlights a worker’s health status as an essential differentiator between user groups. Recognizing the existence of this specific difference between potential end users could provide fresh impetus to more diverse product development that would promote wider adoption of exoskeletons: On the one hand, exoskeletons could focus on the support function, primarily assisting people with an existing health problem (emphasizing rehabilitation). For people without medical issues, on the other hand, comfortable and aesthetically pleasing device design could be prioritized to increase appeal and workers’ intention to use (for prevention). Changes in self-efficacy might also be associated with a user’s physical health: Involving employees with low-back pain, Baltrusch et al. (2021) found self-efficacy enhancement to be strongest for participants who were highly restricted by their medical issues. We have previously shown that the task-specific self-efficacy of healthy workers was reduced at lower levels of perceived physical strain relief and usefulness (Siedl & Mara, 2021). Inferences about interaction effects between exoskeleton use, physical health status and self-efficacy thus require further research.

Finally, focus groups do not account sufficiently for individual differences between participants, for example, in terms of openness to innovation or general opinion regarding exoskeletons, as included in the industrial exoskeleton user acceptance framework established by Elprama et al. (2022). We designed our method to elucidate technology- and workplace-related determinants of exoskeleton acceptance, in the knowledge that there would be differences between individual users that could matter (Venkatesh, 2000). Although the present study follows a qualitative approach with the express purpose of exploring the different ideas, perspectives and representations around the acceptance of occupational exoskeletons, sample size and composition limit generalizability. Hence, caution is advised when interpreting qualitative results in relation to other fields of exoskeleton application. Nevertheless, we are convinced that our findings provide new insights into perceptions of potential exoskeleton users. In conclusion, this work helps to better understand possible barriers to and drivers of exoskeleton use in organizations, and it highlights less-considered variables such as aesthetics and social feedback. The acceptance factors identified not only generate hypotheses for future research, but also offer practical guidance for a more human-centered design and implementation of exoskeletons in the workplace.

Acknowledgments

We thank Anna Katharina Paschmanns for her contribution to discussions about coding and category building, and Prof. Dr. Nicole Kronberger, who supported this work with her expertise in qualitative research methods. Open Access funding provided by Johannes Kepler University Linz.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Austrian Research Promotion Agency (Österreichische Forschungsförderungsgesellschaft) under grant numbers 861519 and 880563.

Notes on contributors

Sandra Maria Siedl

Sandra Maria Siedl is a PhD student in Psychology at Johannes Kepler University Linz and received her master’s degree in Business Administration and Psychology at the Ferdinand Porsche FernFH, Austria, in 2016.

Martina Mara

Martina Mara is a Professor in Robopsychology at Johannes Kepler University Linz, Austria, who obtained her PhD in Psychology at the University of Koblenz-Landau, Germany, in 2014.