A workplace exoskeleton can reduce muscular demand in a controlled test and still fail as a workplace intervention. It can feel impressive during a 20-minute demonstration yet become uncomfortable over a full shift. It can support the intended body region while slowing the task, interfering with movement or shifting strain elsewhere. And it can generate positive worker feedback without ever producing a defensible return on investment.

That is not evidence that occupational exoskeletons do not work. It is evidence that the industry has moved beyond the simpler question of whether wearable assistance can generate force or reduce load under selected conditions.

The harder question in 2026 is whether the right exoskeleton can create repeatable value for the right task, worker population and operating environment.

Recent field evidence makes that distinction unusually clear. An 18-worker masonry study found that one passive back-support exoskeleton increased measured low-back muscular strain by 5.2% during bricklaying, with only 22% of participants willing to use it in the future.^[2] Another study of 23 professional masons found that two passive back-support systems increased mean work-cycle time by 7.77% and 6.28% during a simulated material-handling task.^[3]

At the same time, other field studies show genuine promise. A 24-week randomized trial among logistics workers found that the biomechanical unloading effect of a back-support exoskeleton remained measurable after six months, while remaining users progressively increased their use of the system.^[7] In healthcare, 17 professionals accumulated 312 hours of exoskeleton use over two months, although support and comfort varied substantially between users and tasks.^[6]

The lesson is not that workplace exoskeletons succeed or fail as a category.

They succeed or fail as a specific device-task-worker-deployment combination.

This Insight examines what the latest evidence tells us about those failures. For organisations already preparing an evaluation, our separate practical exoskeleton pilot playbook covers the step-by-step protocol from task assessment through deployment decision.

Key takeaways

• Biomechanical benefit is only one layer of success. A workplace exoskeleton also has to fit the workflow, remain acceptable to workers and make operational and economic sense.
• Lab performance does not reliably predict field acceptance. A 2026 masonry study found that a rigid back-support exoskeleton rated strongly in laboratory testing deteriorated substantially during full-day field exposure.^[4]
• The wrong device-task match can make the target problem worse. In an 18-mason field study, low-back muscular strain increased 5.2% with the tested exoskeleton during bricklaying.^[2]
• Productivity effects need to be measured, not assumed. Two back-support systems increased initial cycle time by 7.77% and 6.28% in a recent masonry experiment, although 22% of individual workers became more productive.^[3]
• Comfort and freedom of movement can matter more for adoption than maximum physical support. Related 2026 construction research found movement hindrance and adjustability more strongly associated with continued-use willingness than physical benefit alone.^[5]
• Short demonstrations systematically miss important information. Heat, contact pressure, workflow interference, transitions, maintenance and declining willingness to wear the device emerge only with realistic exposure.
• ROI remains highly assumption-dependent. A healthcare study calculated that avoiding 18.5 nurse sick days per year could make one implementation cost-neutral, while a different staffing scenario reduced that modelled threshold to 2.7 days. Those were economic scenarios, not observed reductions in absence.^[6]
• International standardisation is moving toward the same process mindset. ISO/DIS 25563 entered the Draft International Standard stage on 24 August 2026 and addresses needs identification, selection, assessment and deployment of workplace exoskeletons.^[10]

Direct answer: why do workplace exoskeleton pilots fail?

Workplace exoskeleton pilots most often fail when an organisation treats the exoskeleton as a product to test rather than an intervention to integrate. The device may provide physical assistance but still be mismatched to the task, unsuitable for part of the workforce, disruptive to normal operations or economically unjustifiable.

The latest evidence suggests that a serious pilot has to pass four different tests:

1. Physical benefit: Does the device reduce the targeted demand without creating an unacceptable new one?
2. Operational compatibility: Can the actual job still be performed safely, efficiently and with acceptable quality?
3. Human adoption: Can representative workers fit, tolerate and voluntarily use the system over realistic periods?
4. Economic viability: Does the measured value justify equipment, implementation, support and operational costs?

A positive result in only one of those four layers is not enough to justify scale.

The 2026 evidence is sending a clear message: task fit beats category hype

A workplace exoskeleton is not an ergonomic effect in itself.

It is a mechanical or powered intervention whose effect changes with posture, movement, assistance setting, fit, environment and the rest of the task.

This explains why studies that appear contradictory can all be valid.

These studies use different devices, tasks, populations and outcome measures and should not be pooled as if they were equivalent experiments. The table is intended to show recurring implementation patterns, not rank the systems tested.

Failure mode 1: selecting the exoskeleton before defining the problem

One of the easiest ways to design a weak pilot is to begin with a device.

A supplier arrives. Workers try it. Management asks whether they like it. A few people report that lifting feels easier. The project is then called a pilot.

That reverses the decision process.

The organisation should first define the exposure it is trying to change:

• repeated shoulder elevation;
• sustained overhead tool use;
• frequent trunk flexion;
• floor-level lifting;
• static forward-bent posture;
• repeated knee loading;
• load carriage;
• another measurable physical demand.

The intended assistance mechanism then has to match that exposure.

This principle matters because jobs described by a single label can contain completely different movement demands. “Masonry”, “warehouse work” and “assembly” are not individual tasks.

A mason may lift, carry, rotate, bend, kneel, align blocks, apply mortar and walk across an uneven site. A warehouse operator may lift successfully with a back-support system and then immediately need to walk, sit in a vehicle, climb a step or rotate in a restricted aisle.

That helps explain the negative 2026 masonry result. The authors concluded that the tested design had limited suitability for highly dynamic, multi-planar bricklaying work. The low-back effect was not merely neutral: measured strain increased 5.2%.^[2]

The right conclusion is not “back exoskeletons increase back strain.” The study tested one device-task combination.

The more useful conclusion is that the label on the product does not guarantee that its assistance matches every task involving the same body region.

Failure mode 2: mistaking a laboratory win for a workplace win

Laboratory evaluation is valuable. It makes movements repeatable, enables instrumentation and allows researchers to isolate mechanisms.

But the laboratory deliberately removes many of the variables that determine whether a worker will continue using an exoskeleton.

A 2026 construction study provides an unusually clear example.

Researchers compared a rigid HAPO Back system with the softer BISKO system across laboratory and field masonry conditions. Twenty-three masons participated in the laboratory component and 15 in field testing.^[4]

On the study’s five-point scales, where lower scores generally represented more favourable assessments, HAPO’s median overall rating deteriorated from 1.5 in the laboratory to 3.0 in the field. Its median willingness-to-continue-use rating moved from 1.0 in laboratory testing to 4.0 in the field. The soft BISKO system remained much more stable between environments.^[4]

That is exactly the type of information a 15-minute demonstration cannot reveal.

Failure mode 3: optimizing one metric while breaking another

Exoskeleton studies frequently measure muscle activity because electromyography can detect whether muscular demand changes during a task.

That is valuable evidence.

It is not a complete deployment decision.

A device could reduce targeted muscle activity while simultaneously:

• increasing task time;
• changing movement strategy;
• increasing pressure elsewhere;
• restricting walking or rotation;
• reducing precision;
• increasing heat;
• interfering with tools or PPE;
• being worn so infrequently that the biomechanical effect becomes operationally irrelevant.

The recent masonry cycle-time study illustrates this trade-off quantitatively.

Twenty-three professional masons moved 17.2 kg concrete masonry units under three conditions: no exoskeleton, the soft BISKO back-support system and the rigid HAPO Back system.^[3]

Average baseline cycle time was 8.11 seconds. That increased to 8.74 seconds with BISKO and 8.62 seconds with HAPO Back, corresponding to increases of 7.77% and 6.28%. The difference between the two exoskeletons was not statistically significant.^[3]

But the individual results make the study more interesting than the averages: 22% of participants demonstrated productivity improvements, and older and more experienced workers tended to perform more favourably with exoskeleton use.^[3]

That is not evidence for assigning exoskeletons based on age. The sample was small and all participants were male. It does show why an organisation should preserve participant-level data rather than reporting only the mean.

A 7% initial cycle-time penalty may also be acceptable in one operation and unacceptable in another. If the intervention meaningfully reduces fatigue across a difficult task, workers may adapt with experience or the business may accept a small throughput cost for another benefit.

The correct question is not simply:

“Did productivity fall?”

It is:

“What changed physically, operationally and behaviourally, by how much, for which workers, and is that trade-off acceptable for this operation?”

Failure mode 4: treating worker acceptance as a satisfaction survey

“Would you use this exoskeleton?” is a useful question.

It is not an adoption metric.

Actual adoption is behaviour.

A stronger workplace evaluation records:

• how many eligible workers actually wear the device;
• how often they wear it;
• how many minutes or hours it is worn;
• which tasks trigger use;
• which tasks trigger removal;
• how often workers adjust it;
• who stops using it;
• when they stop;
• why they stop;
• whether use increases or declines after familiarisation.

This distinction appears repeatedly in field research.

In the 2025 randomized logistics trial, the back-support exoskeleton continued to reduce back-muscle activity during standardized lifting after 24 weeks, and the workers who remained in the intervention progressively increased their daily use. But three exoskeleton users stopped using the device during the intervention period.^[7]

An earlier five-week logistics field study produced another instructive result. A passive shoulder exoskeleton reduced anterior-deltoid muscle activity by approximately 13–39% and upper-trapezius activity by 16–60% during measured tasks, yet adherence to the familiarisation protocol was too low to support the intended conclusions and workers’ perceptions of using the device deteriorated over time.^[8]

That is a particularly important industry lesson:

A product can be biomechanically effective when it is worn and still fail as a workplace intervention because workers do not keep wearing it.

Comfort is not a cosmetic metric

Comfort is sometimes treated as secondary to engineering performance.

For wearable technology, that distinction is artificial.

If discomfort causes removal, the effective assistance delivered over the workday moves toward zero.

Related 2026 masonry research compared a rigid HAPO Back system with a soft textile BISKO design across 15 masons on five construction sites. The researchers reported greater perceived reductions in fatigue and exertion with BISKO and substantially better ratings for comfort, adjustability and movement freedom.^[5]

Most notably, movement hindrance and adjustability were more strongly correlated with willingness to continue using the exoskeleton than physical benefit alone: the reported correlation coefficients were r = 0.79 for movement hindrance and r = 0.76 for adjustability.^[5]

The study was small and based partly on subjective assessment, so those coefficients should not be generalized to every workplace.

But the direction is commercially important.

Maximum assistance is not automatically the design target that produces maximum adoption.

Failure mode 5: testing the “average worker”

There is no average workforce.

Height, body proportions, sex, strength, previous symptoms, clothing, work technique, experience and individual preference can all affect how an exoskeleton interacts with its user.

A 2026 systematic review examined demographic representation across 191 arm- and back-support exoskeleton studies published between 2019 and 2024.^[9]

Approximately 78% of participants were male. None of the studies reported race or ethnicity, and anthropometric diversity was limited. Only 25 of the 191 studies incorporated demographic variables into statistical analysis.^[9]

The review found moderate evidence that females may experience larger reductions in trunk-extensor activity with back-support exoskeletons, but evidence for most other demographic differences remained insufficient.

The important procurement lesson is not that one demographic group should receive a particular device.

It is that research performed on narrow participant groups cannot automatically prove fit or effectiveness across a diverse workforce.

An internal pilot should therefore report people who cannot fit the device as carefully as people who can.

Excluding difficult-to-fit employees from the analysis may make the product appear successful while hiding a deployment problem.

What the 2026 adoption review tells us about the real barriers

The broadest recent synthesis comes from a systematic review published in the Journal of Occupational Rehabilitation in May 2026.

Researchers reviewed 28 studies and identified 70 unique facilitators and 67 unique barriers to occupational-exoskeleton adoption.^[1]

The distribution is revealing.

The review also found that studies involving up to one week of actual exoskeleton exposure identified fewer unique barriers than studies where participants had no prior hands-on experience: 25 versus 50.^[1]

That supports the value of familiarisation.

But it should not be misread as proof that barriers simply disappear with use. The authors explicitly identify the lack of long-term exposure evidence as an important limitation, and the individual studies above demonstrate that some problems emerge only after longer or more realistic exposure.

The systematic review itself also assessed the overall risk of bias as moderate, with recurring issues including small or convenience samples and limited reporting.^[1]

So even in 2026, the evidence does not support universal deployment rules.

The Exoskeleton Index Four-Layer Pilot Test

Based on the evidence above, we think organisations should stop asking whether a pilot was simply “successful” and instead ask whether it passed four separate layers.

A pilot does not need every individual metric to improve.

It does need a predefined explanation for which trade-offs are acceptable.

For example, an organisation might accept a modest short-term cycle-time penalty if the device substantially reduces a known physical exposure and users become faster after familiarisation. It should not retrospectively invent that justification after observing disappointing productivity data.

Which tasks are stronger candidates for an exoskeleton pilot?

No table can determine whether a workplace exoskeleton is safe or suitable for an individual operation.

But the evidence does support useful screening principles.

This is Exoskeleton Index screening guidance, not a safety classification or substitute for a task-specific risk assessment.

For detailed product-selection criteria, evidence review and supplier evaluation, see our practical exoskeleton buyer’s guide and back-support exoskeleton buyer’s guide.

Exoskeletons should not be used to preserve avoidably bad work design

There is another failure mode before the pilot even begins: selecting a wearable intervention when the hazard can be more reliably removed or redesigned.

NIOSH has cautioned that occupational exoskeletons should be considered for residual risks that cannot feasibly be eliminated through engineering controls rather than as a reason to skip higher-level risk reduction.^[11]

NIOSH also identifies possible risks including contact pressure, reduced mobility, changes in centre of gravity and risk transfer. If an exoskeleton allows a worker to hold a tool for longer, for example, exposure to vibration, noise or airborne hazards could theoretically increase unless those exposures are also controlled.^[11]

The question before a pilot should therefore be:

“Why is wearable assistance the right intervention for this residual demand?”

If workstation redesign, a manipulator, tool balancer, conveyor, lift table or process change can remove the exposure more reliably, that alternative belongs in the comparison.

Failure mode 6: assuming ROI instead of modelling it

Return on investment is one of the weakest areas in exoskeleton decision-making because the most commercially attractive outcomes are also among the hardest to prove quickly.

A short pilot may measure:

• muscle activity;
• perceived exertion;
• discomfort;
• cycle time;
• quality;
• wearing time.

It generally cannot prove that a company will experience fewer musculoskeletal injuries over the following five years.

Those are different levels of evidence.

The healthcare example shows how sensitive ROI is to assumptions

The 2026 University Hospital Magdeburg field study provides a useful example because the researchers explicitly modelled economic viability.

Seventeen healthcare professionals used a passive exoskeleton over two months, accumulating 312 total hours of wear, approximately 20 hours per participant across three shifts. All participants considered the system easy to use, but ratings for support and comfort varied, and the average overall German school grade was 2.7 ± 0.99, approximately “satisfactory”.^[6]

The paper calculated that the investment could become cost-neutral if at least 18.5 nurse sick days per year were prevented. When indirect effects associated with staff shortages were incorporated, the threshold in the model fell to 2.7 days.^[6]

Those numbers are valuable precisely because they show how much the answer changes when the economic assumptions change.

The study did not observe that the exoskeleton prevented 18.5 or 2.7 sick days. These are break-even scenarios.

A new 2026 construction model is more bullish, but it is still a model

On 27 August 2026, researchers at Louisiana State University published an agent-based simulation of exoskeleton adoption in construction through the Construction Research Congress.^[12]

Under modelled strategic-adoption scenarios, the simulation reported:

• total project-cost reductions of up to 6.7%;
• project-duration reductions of up to 7.8%;
• approximately 45% reductions in modelled health-related costs and injury-recovery time;
• ROI as high as 90% in large-scale project scenarios.

Those results are interesting and unusually timely.

They are simulation outputs, not observed financial performance from completed exoskeleton deployments. Their usefulness is in showing which assumptions and system interactions can make an investment economically attractive, not in establishing that construction employers should expect a 90% return.

The contrast between the healthcare break-even model and the new construction simulation makes the core point:

There is no universal exoskeleton ROI.

A better way to calculate workplace exoskeleton ROI

An organisation should build a task-specific economic model from the evidence generated in its own evaluation.

A useful starting structure is:

Each assumption should then be shown under at least three scenarios:

• Conservative: little or no health-cost benefit and observed pilot productivity impact retained.
• Base case: most defensible assumptions supported by the pilot and organisation’s historical data.
• Upside: plausible but not yet demonstrated reductions in absence, overtime or other downstream costs.

This prevents an attractive ROI percentage from hiding a speculative injury-reduction assumption.

Do not force every worker to use the same device in the same way

The 23-mason productivity study found that average cycle time worsened while 22% of individual participants improved.^[3]

The healthcare study found that participants largely selected particular tasks for exoskeleton use rather than wearing the device indiscriminately through every activity.^[6]

These results suggest that the most mature deployment model may not always be:

“Every eligible worker wears Device X for the entire shift.”

It may instead be:

“The device is available to workers who fit it, for the defined tasks where evidence shows sufficient benefit, under conditions where interference remains acceptable.”

Selective deployment is not necessarily evidence that the technology failed.

It may be evidence that the organisation understood where it actually creates value.

When should an exoskeleton pilot stop?

This is one of the most important questions in a pilot and one of the least attractive for a supplier to discuss.

A pilot should not continue merely because equipment has been purchased, management has publicly supported the project or significant employee time has already been invested.

A failed pilot is not necessarily a failed technology

This distinction matters for both buyers and manufacturers.

If an exoskeleton performs poorly during masonry work, that does not prove it will perform poorly during repetitive warehouse lifting.

If workers reject a rigid design, they may accept a softer architecture.

If full-shift use is impractical, task-specific use may still make sense.

If one user group cannot achieve appropriate fit, another product may cover the workforce more effectively.

The purpose of a pilot should therefore be to distinguish between at least four possible outcomes:

1. Stop: the intervention does not create sufficient value or introduces unacceptable risk.
2. Modify and retest: the task, fitting, training, configuration or device selection needs to change.
3. Deploy narrowly: value exists for a defined worker-task combination but not for universal use.
4. Scale with monitoring: evidence supports broader deployment while longer-term outcomes continue to be tracked.

Our full pilot playbook provides the detailed gated process for reaching those decisions.

ISO 25563 has just moved forward, and that matters

One of the clearest signals that occupational exoskeletons are becoming a more mature technology category arrived only days before this article was prepared.

On 24 August 2026, ISO registered ISO/DIS 25563 at Draft International Standard stage 40.00. The draft is now in the ISO member-enquiry phase, with a 12-week DIS ballot.^[10]

Its title is:

Ergonomics — Process for the integration of wearable physical assistive devices (exoskeletons) — Expression of needs, selection, design, assessment and deployment.

The draft addresses workplace use across sectors and organisation sizes and is intended for user organisations as well as designers, manufacturers and integrators. It excludes exoskeletons for medical, rehabilitation, games, sport and leisure applications.^[10]

Two boundaries are important:

• ISO/DIS 25563 is not yet a published International Standard.
• It does not provide exoskeleton certification.

We should therefore not describe a product or deployment as “ISO 25563 certified”.

What matters today is the direction of travel.

The emerging international framework is structured around a process: identify the need, select, assess, integrate and deploy.

That closely reflects what the field evidence is showing.

What a serious workplace exoskeleton evaluation should report

The industry will become easier to evaluate when companies publish more than before-and-after testimonials.

A useful workplace report should identify, at minimum:

The more consistently the industry publishes these fields, the easier it becomes for buyers to distinguish genuine deployment evidence from a successful demonstration.

What buyers should ask vendors before a pilot

For organisations already screening suppliers, the most useful questions are often not the headline specifications.

• Which exact tasks was this model designed to assist?
• Which tasks should it not be used for?
• What independent field evidence exists for work similar to ours?
• How long were participants exposed in those studies?
• What proportion of workers stopped using the device?
• What fit range has actually been validated?
• How does the system behave during walking, sitting, turning, kneeling and other non-target movements?
• What negative or neutral findings have been reported?
• What PPE or tool conflicts are known?
• What training and fitting support will the supplier provide?
• What components wear out and what is their replacement cost?
• Where is the device repaired?
• What uptime and spare-parts support can be expected?
• For powered devices, what battery and software-management burden should be included?
• Will the supplier support an evaluation whose outcome may be not to purchase?

A serious supplier should be able to discuss the situations where its product does not fit.

The business case is becoming more demanding, not less attractive

It would be easy to interpret mixed field findings as bad news for occupational exoskeletons.

We see the opposite.

The category is reaching the stage where simplistic claims are being replaced by harder questions about repeatable value.

That is what happens when a technology moves closer to normal procurement.

Manufacturers are being asked to demonstrate:

• task-level effectiveness;
• fit across real workforces;
• sustained acceptance;
• integration with existing equipment;
• service and maintenance capability;
• transparent evidence;
• economic value under realistic assumptions.

The companies that can answer those questions have a stronger commercial proposition than companies that compete only on maximum support, actuator power or a dramatic demonstration.

This also changes what buyers should expect from the supplier relationship.

The winning business model may increasingly include task assessment, fitting, training, pilot support, maintenance, data interpretation and ongoing deployment services alongside the physical device.

That would make occupational exoskeletons look less like stand-alone equipment and more like integrated ergonomic systems.

Where workplace exoskeleton adoption looks strongest

The evidence does not justify claiming that one industry will universally adopt exoskeletons first.

But the strongest deployment opportunities share characteristics.

They involve physical demands that are:

• meaningful enough to justify intervention;
• difficult to eliminate fully through fixed engineering controls;
• frequent or sustained enough for assistance to matter;
• predictable enough for the device to support without constant interference;
• performed by workers the product can reliably fit;
• located in an environment where service, hygiene, PPE and workflow can be managed.

This is why manufacturing and assembly, logistics and warehousing and selected construction and field-operation tasks remain strategically important parts of the occupational exoskeleton market.

But even inside those industries, adoption should happen task by task rather than sector by sector.

Exoskeleton Index position: the next phase of the market is integration

We remain bullish on the long-term occupational exoskeleton category.

Not because every field trial is positive.

Because the evidence is becoming detailed enough to show where wearable assistance creates value, where it fails and what must improve.

The market’s early phase rewarded proof that a wearable mechanism could reduce physical demand.

The next phase will reward repeatability:

• Can the supplier identify the right task?
• Can the device fit a representative workforce?
• Can users wear it for long enough to matter?
• Can it coexist with the rest of the job?
• Can the organisation support it operationally?
• Can the economics survive conservative assumptions?
• Can the result be replicated across sites?

That is a higher bar.

It is also how the category becomes credible enough for larger-scale procurement.

The latest research suggests that the most successful implementations will not be the companies that simply buy an exoskeleton and ask workers to wear it. They will be the organisations that treat wearable robotics as one component inside a broader ergonomics, operations and workforce system.

The best pilot may end with “no”

The purpose of an exoskeleton pilot is not to prove that buying the exoskeleton was a good idea.

It is to discover whether the intervention creates enough value to justify continuing.

A pilot that establishes that a device:

• does not fit the task;
• cannot accommodate enough workers;
• creates unacceptable workflow interference;
• shifts strain elsewhere;
• is not worn consistently;
• cannot generate a credible business case;

has produced useful evidence.

The failure would be scaling first and discovering those facts later.

For organisations moving from this evidence review into implementation, use our Exoskeleton Pilot Playbook to build the protocol, browse the Product Directory to identify candidates, or use Compare to examine products side by side.

Methodology

This Insight was researched and reviewed using public information available through 29 August 2026.

Its purpose is to examine why occupational exoskeleton evaluations and deployments underperform despite evidence of physical assistance, and to translate recent research into practical decision principles for employers, safety teams, ergonomists, procurement professionals, researchers and exoskeleton suppliers.

We prioritised:

1. peer-reviewed systematic reviews;
2. recent real-world field studies;
3. controlled studies relevant to productivity, biomechanics and adoption;
4. longer-duration occupational trials;
5. primary standards information from ISO;
6. official occupational-safety guidance from NIOSH;
7. recent economic research where the modelling assumptions and evidence type could be clearly distinguished.

The May 2026 systematic review of occupational-exoskeleton adoption included research published through 2025. We therefore supplemented it with field studies and economic research published during 2026 rather than assuming that the systematic review represented the newest available evidence.

We did not combine numerical outcomes across studies because participants, devices, tasks, exposure periods and measurement methods differ substantially.

Manufacturer claims were not used as proof of injury prevention or economic benefit.

How to interpret the evidence

A reduction in muscle activity or perceived exertion supports a claim about the measured condition. It does not by itself prove a reduction in injuries, absence or workers’ compensation costs.

Similarly, a productivity change during a short controlled task should not automatically be extrapolated to long-term production.

Economic simulations and break-even analyses are identified as models rather than observed financial outcomes.

The Exoskeleton Index Four-Layer Pilot Test and task-screening table are editorial decision frameworks created for this article. They are not regulatory standards, validated ergonomic instruments or substitutes for professional risk assessment.

Important limitations

The occupational exoskeleton literature still contains many small samples, short exposure periods and laboratory studies. Device generations also change faster than long-term health outcomes can be established.

The strongest conclusion supported by the current evidence is therefore not that occupational exoskeletons universally prevent injury or generate ROI.

It is that their value is highly dependent on the interaction between device, task, worker and deployment process, and that organisations can substantially improve decision quality by measuring those dimensions separately.

References

1. van Laar DM, Bieleman HJ, Rijnders CSI, et al. Occupational Exoskeletons as a Technological Solution for Musculoskeletal Disorders: A Systematic Review of Barriers and Facilitators for Sustainable Workplace Integration. Journal of Occupational Rehabilitation. Published 7 May 2026. DOI: 10.1007/s10926-026-10397-5.
2. Brandt M, et al. When technology misaligns with work demands: a biomechanical and ergonomic field study of exoskeleton use in masonry. Applied Ergonomics. 2026;135:104767. DOI: 10.1016/j.apergo.2026.104767.
3. Al-Khiami MI, Lindhard SM. EMG-Derived Cycle-Time Assessment of Passive Back-Support Exoskeletons in Construction Workers During Repetitive Masonry Manual Handling Tasks. International Journal of Construction Management. Published online 7 August 2026. DOI: 10.1080/15623599.2026.2711934.
4. Al-Khiami MI, Lindhard SM, Saad AS. Laboratory and field evaluation of user-perceived effort, comfort, and acceptance of passive back-support exoskeletons for masons. International Journal of Industrial Ergonomics. 2026;112:103881. DOI: 10.1016/j.ergon.2026.103881.
5. Al-Khiami MI, Lindhard SM. Field comparison of rigid and soft passive back-support exoskeletons: physical exertion, comfort, and adoption among masonry workers. Safety Science. 2026;197:107138. DOI: 10.1016/j.ssci.2026.107138.
6. Exner B, Waßmann S, Gück D, Frielitz F. Passive exoskeletons in healthcare practice: Usability and acceptance in a clinical setting. Applied Ergonomics. 2026;135:104746. DOI: 10.1016/j.apergo.2026.104746.
7. Jakobsen LS, et al. Effects of 24-weeks in-field use of a back-supporting exoskeleton on biomechanics, work intensity and musculoskeletal discomfort: A randomized controlled trial among logistic workers. Applied Ergonomics. 2025;125:104469. DOI: 10.1016/j.apergo.2025.104469.
8. Jakobsen LS, de Zee M, Samani A, Desbrosses K, Madeleine P. Biomechanical changes, acceptance, and usability of a passive shoulder exoskeleton in manual material handling: A field study. Applied Ergonomics. 2023;113:104104. DOI: 10.1016/j.apergo.2023.104104.
9. Kazemi Z, Park J, Srinivasan D. Unequal assistance? A systematic review of demographic representation and outcome differences in exoskeleton studies. Applied Ergonomics. 2026;137:104808. DOI: 10.1016/j.apergo.2026.104808.
10. International Organization for Standardization. ISO/DIS 25563 — Ergonomics — Process for the integration of wearable physical assistive devices (exoskeletons) — Expression of needs, selection, design, assessment and deployment. Draft International Standard registered 24 August 2026.
11. National Institute for Occupational Safety and Health. Industrial Exoskeletons. NIOSH Science Bulletin. U.S. Centers for Disease Control and Prevention.
12. Soleymani M, Bonyani M, Wang C. Quantifying the Economic Viability of Adoption of Exoskeleton Technology in Construction via Agent-Based Simulation. Construction Research Congress 2026. Published online 27 August 2026. DOI: 10.1061/9780784486979.091.

Editorial note: Exoskeleton Index is bullish on the long-term potential of occupational wearable robotics but does not treat exoskeletons as universal injury-prevention devices. Positive biomechanical findings, user preferences, productivity effects and economic models answer different questions and should not be presented as interchangeable evidence. Product suitability should be assessed for the specific worker, task, environment and jurisdiction.