Wearables in OSH

Types of wearables and preventative applications

Exposure wearables measure physical and chemical agents in the environment directly on the wearer: noise dosimeters, personal gas and oxygen detectors, temperature and humidity sensors for heat stress, radiation dosimeters, and accelerometers for vibrations. Their advantage over ambient measurements is that they record each person’s actual exposure throughout the day and allow for immediate alerts when thresholds are exceeded, although they do not replace the standardized measurements required by specific regulations for exposure assessment.

Safety and location wearables include proximity tags and beacons that alert workers and drivers when someone approaches a vehicle or machine; dead man’s switch and fall detection devices for solo work, with automatic notification to a control center; and indoor location systems for emergencies, confined spaces, and evacuations. Ergonomic and physiological wearables record postures, movements, exertion, and, in some cases, heart rate or body temperature for analyzing physical workload, fatigue, and heat stress. Active or passive exoskeletons assist users in handling tasks or working with their arms raised and can integrate sensors.

Smart personal protective equipment integrates sensors into helmets, vests, footwear, or gloves to verify use, detect impacts or exposures, and communicate alerts. When the device serves a protective function, it must comply with Regulation (EU) 2016/425, and the electronic integration cannot reduce its protective performance.

Benefits, limitations and risks

• Real-time detection. Immediate alerts for exposures or risk situations that periodic measurements do not detect.
• Individual exposure. Data on each person’s actual exposure, useful for adapting tasks, breaks, and rotations.
• Vulnerable groups. Support for single workers, newly hired workers, older workers, or workers with disabilities.
• Data for prevention. Aggregated indicators to review risk assessment and prioritize measures.
• Accuracy and reliability. Consumer sensors do not always achieve the accuracy of standardized measuring equipment; the data must be validated and do not replace measurements required by regulations.
• Privacy and surveillance. The continuous collection of personal data, especially physiological and location data, generates risks of surveillance, misuse of data, and loss of trust.
• Psychosocial risks. The feeling of control, pressure to perform, and dependence on alerts can generate stress and reduce attention to risk.
• Device ergonomics. Weight, grip, battery life, compatibility with protective equipment, and comfort all influence actual use.
• Safety and maintenance. Calibration, batteries, upgrades, cybersecurity, and replacement.

Legal and governance requirements

• Risk assessment and consultation. The introduction of wearables is a new technology that requires consultation with worker safety representatives and a review of the assessment, including psychosocial risks.
• Data protection. Define purpose, lawful basis, minimisation, retention and access, and inform individuals. Physiological data are health data where they reveal information about health status under GDPR Article 4(15); a valid Article 9 condition is then needed in addition to an Article 6 basis. A prior impact assessment is required where processing is likely to result in high risk, including the cases and lists under Article 35; simply calling any monitoring systematic does not settle the assessment. In Spanish occupational health surveillance, personal medical information has the safeguards of Law 31/1995 Article 22, distinct from preventive conclusions the employer may receive.
• Digital rights. Article 20 bis of the Workers’ Statute and articles 87 to 91 of Organic Law 3/2018 on privacy in relation to digital devices, geolocation and video surveillance.
• Artificial intelligence regulations. Classification depends on intended purpose and the criteria in Article 6 and Annex III of Regulation (EU) 2024/1689; point 4(b) includes specified uses for monitoring and evaluating work performance or behaviour. Check the Article 6(3) exceptions, their limits and the applicable timetable, rather than automatically treating every sensor as high-risk AI. Workplace emotion inference is prohibited except for the medical or safety use described in Article 5(1)(f), which does not remove other data protection duties.
• Personal protective equipment. Compliance with Regulation (EU) 2016/425 when the wearable is part of protective equipment.
• Purpose, safeguards and non-discrimination. Distinguish a voluntary wellbeing programme from equipment required for a task. Data collection needs justification and proportionality; accepting a device does not itself legitimise every form of processing. Prevent uses incompatible with the preventive purpose and provide alternatives or adjustments where needed. Occupational health surveillance data must not be used discriminatorily or to the worker’s detriment.
• Participation. Design, testing and evaluation with users and their representatives, as recommended by EU-OSHA.

Organizational application: how to implement wearables successfully

1. Define the specific preventive purpose (exposure, proximity, working alone, ergonomics) based on the risk assessment and verify that the previous technical and organizational measures are implemented.
2. Consult the project with the worker safety representatives and carry out the data protection analysis: purpose, minimization, retention, impact assessment, separation of health data and, if applicable, analysis in accordance with Regulation (EU) 2024/1689.
3. Select devices with validated accuracy, suitable ergonomics, compatibility with protective equipment and compliance with Regulation (EU) 2016/425 when they are part of protective equipment.
4. Conduct a pilot test with users, measure reliability, false positives and acceptance, and adjust thresholds, alerts and response procedures.
5. Clearly inform staff about measurements, purpose, access and retention, and establish controls preventing uses incompatible with the preventive programme or contrary to occupational health surveillance safeguards.
6. Integrate alerts and aggregated data into the preventive management system, with alert response, assessment review, and protocols for solo work and emergencies.
7. Maintain the devices (calibration, batteries, updates, cybersecurity) and periodically review the effectiveness, psychosocial effects and regulatory compliance.

Preventive management software allows the integration of alerts and aggregated data from wearables with risk assessment, exposed positions and people, management of resulting actions and solo work protocols, and maintenance of traceability, with separation of health data and recording of information and consultations carried out.

Limits and common mistakes

1. Replacing standardized exposure measurements with unvalidated consumer sensor data.
2. Implementing wearables without consulting representatives, without data protection analysis, or providing clear information to staff.
3. Processing personal signals without a lawful basis and safeguards, disclosing occupational health surveillance medical information, or reusing data for an incompatible purpose.
4. Relying on alerts without a response procedure or human supervision, especially in solo work.
5. Ignoring the ergonomics of the device and its compatibility with protective equipment leads to it not being used.
6. Failure to maintain or calibrate devices or review the psychosocial effects of continuous monitoring.

Specific data protection, artificial intelligence and protective equipment requirements must be analyzed on a case-by-case basis; this sheet is for informational purposes only.

Regulatory and reference framework

• Law 31/1995, of November 8. Law on the Prevention of Occupational Risks; risk assessment and consultation regarding new technologies.
• Regulation (EU) 2016/679 . General Data Protection Regulation; health data, impact assessment and processing principles.
• Organic Law 3/2018, of December 5. Protection of Personal Data and guarantee of digital rights; digital rights in the workplace.
• Regulation (EU) 2024/1689 . Artificial Intelligence Regulation; high-risk systems and prohibited practices in the workplace.
• Regulation (EU) 2016/425 . Personal protective equipment; conformity of smart equipment with protective function.
• EU-OSHA. Smart digital monitoring systems for occupational safety and health: uses and challenges (2022) . Report on uses, opportunities and challenges of smart monitoring systems, including wearables.
• EU-OSHA. Digitalization of work . Thematic portal on intelligent safety and health monitoring systems.

Royal Decree 773/1997 regulates the use of personal protective equipment in the company, and the specific standards for noise, vibrations, chemical agents and thermal stress establish the standardized measurement methods that wearables complement.

Mining and confined spaces: what a signal can establish

In mining and confined spaces, distinguish three questions: where a person is, what conditions exist around them and what physiological response is observed. A heart-rate wristband does not establish that the atmosphere is breathable; a location tag does not establish that someone is conscious. The programme must retain the measurements, authorisations, external attendance and rescue preparation appropriate to the task. Connectivity does not authorise entry or turn a hazardous intervention into safe lone working.

For each signal, document the variable, sensor location, validated range, update frequency and meaning of missing data. Trials should include depleted batteries, accidental removal and connection loss, with a defined response for each. Physiological data require assessment of whether health information is revealed and application of the corresponding safeguards; they must not become automatic medical fitness decisions. Equipment and performance must be verified for the particular environment, including special conditions where present.

Proximity and indoor location: testing the complete warning chain

A proximity alert and an indoor positioning system solve different problems. The former may warn of interaction with machinery; the latter locates a tag relative to a facility. Before deployment, check coverage, obstructions, position updates and operation with actual clothing and equipment configurations. Testing must establish that the warning reaches the person expected to act and that a safe response is possible, without exposing people to a test collision.

A 2018 NIOSH study documented interference between particular personal dust monitors and mining proximity systems. Its results do not establish a universal separation distance for all devices: they demonstrate why compatibility of the complete arrangement needs verification. During an emergency, the last transmitted position may be outdated, and the interface should show its age. Location data complement headcounts and communication; they neither replace physical separation measures nor independently establish that evacuation is complete.

Sensors in PPE: material damage and chemical protection

A piezoelectric signal may represent loading measured by a particular system; inferring accumulated material fatigue requires a validated relationship with that equipment and its use. An impact record must not automatically become a remaining-service-life calculation. For PPE, retain inspections, withdrawal criteria and manufacturer instructions, and check conformity of the combination when integrating components. Electronics must not introduce snagging hazards or impair fit, cleaning or protection.

For textiles containing chemical sensors used in HAZMAT operations, identify which substances they detect and under what conditions. Detecting a substance does not demonstrate the garment’s resistance to permeation or protection against other substances present. Ensemble selection still depends on the assessed hazard. Sensor data may complement removal and decontamination arrangements but do not justify remaining exposed when adequate protection is absent. Distinguish an investigated prototype from a documented capability of the purchased product.

Postural biofeedback and pressure inside footwear

Postural feedback requires defining the movement recognised by the device and checking it during normal tasks. A position may be necessary for a particular operation; repeated warnings without changes to working height, access or pace shift the problem onto the individual. A pilot should record detection errors, discomfort and actual exposure changes with user participation. Fewer alerts do not independently demonstrate fewer injuries.

Instrumented insoles provide pressure or support data within their validated scope. They do not automatically measure the entire musculoskeletal load or diagnose injury. Compare findings with task observation, footwear and floor conditions, and check that the insole does not alter the fit or protective performance of safety footwear. Analysis should support redesign of loads and work organisation, rather than rank individual performance against an allegedly ideal gait.

Sweat and hydration: separating validation from health benefit

Research exists on microfluidic sensors that record sweat and provide wearer alerts. A 2025 study evaluated a system using local measurements and estimated whole-body losses; several authors declared links to the developing company. This is evidence about that device and its tests, not validation of every wristwatch or automatic proof of heatstroke prevention. The body sample, estimation model and conditions of use must be identified.

Before using a result to decide measures, review comparators, errors, study population and the need for independent verification. An alert does not replace access to water, work organisation, rest and environmental control. A sweat reading must not become an individual fluid or electrolyte prescription without professional assessment. System acceptance can examine data coverage, failures, alert response and effectiveness of measures, while keeping technical validation distinct from health outcomes.