What is preventative automation?
The expression describes a preventative goal, not a specific technology. It can encompass automatic barriers and interlocks, safe shutdown systems, sensor-activated ventilation, automated guided vehicles, robots that handle hazardous materials, automatic alerts, or digital workflows that ensure no measure goes unmonitored. The value lies not in automation itself, but in better controlling an identified risk.
Law 31/1995 mandates, first and foremost, the prevention of risks, combating them at their source, and adapting work to the individual. Therefore, replacing hazardous exposure with automation can occupy a more robust position in the prevention hierarchy than simply issuing warnings or providing personal protective equipment (PPE). However, this solution can introduce entrapment hazards, unexpected movements, sensor failures, mental overload, loss of autonomy, or psychosocial risks. The outcome must be assessed in relation to the entire work system: people, tasks, organization, environment, and technology.
What tasks can be automated?
EU-OSHA distinguishes between the automation of physical and cognitive tasks. Physical tasks include handling loads, inspecting hazardous areas, palletizing, welding, and internal transport. Cognitive tasks can include detecting deviations, classifying alerts, and verifying that documentation steps have been completed. In both cases, it is advisable to automate specific tasks and not assign a generic preventive responsibility to technology.
Some functions must remain under professional judgment: interpreting a complex assessment, accepting residual risk , determining fitness for health, or deciding on measures that significantly affect people. A system can alert, prioritize, or block a defined condition, but this does not replace the employer’s duty of care or the competence of preventative personnel. Human intervention must be planned: who supervises, when the system can be stopped, what happens in the event of a failure, and how safe operation is restored.
Differences with related concepts
- Digitization of occupational risk prevention: converts information and processes into digital formats; it can exist without automatically executing actions.
- Industrial automation: seeks to perform tasks with less human intervention; it is only preventive when its design aims to control risks and is integrated into the management of occupational health and safety .
- Artificial intelligence applied to occupational safety and health: uses models capable of inferring, classifying, or generating results. Automation can function with fixed rules and without AI.
- Sensors applied to occupational safety and health (OSH): they capture variables from the environment, equipment, or process. Measuring is not the same as controlling: it is necessary to define thresholds, response times, maintenance procedures, and responsible parties.
- OSH/OHS software: organizes information and workflows; automating reminders or evidence does not replace verifying the actual effectiveness of the measures.
The decisive difference is functional: a tool is preventive if it is part of a planned measure against a risk and if its effectiveness can be verified.
Benefits and risks that need to be evaluated
Automation can remove people from hazardous environments, reduce manual handling, repetitive movements , and cognitive load, and stabilize controls that depend on a precise sequence. It can also facilitate early detection, automatic shutdown, and traceability. EU-OSHA warns, however, of physical, psychosocial, and organizational effects: over-reliance, automation bias, low acceptance, fear of job loss, machine-imposed pace, reduced autonomy, and new supervisory demands.
Both normal operation and foreseeable failures, cleaning, adjustment, maintenance, restarts, manual intervention, and the coexistence of people and equipment must be evaluated. In connected systems, a cybersecurity incident can have physical safety consequences. Preventive measures are demonstrated through data and observation of actual work; they cannot be assumed based on the novelty of the technology.
How it is applied in practice
- Define the problem to be prevented. Identify the hazard, the people exposed, and the desired outcome.
- Evaluate before change. Compare alternatives and prioritize elimination, substitution, and collective protection .
- Consult with the workers. Law 31/1995 requires prior consultation on new technologies with effects on safety and health.
- Design the work system. Establish zones, interfaces, limits, stop, recovery, supervision and responsibilities.
- Validate and verify. Check installation, safety functions and fault response before commissioning.
- Inform and train. Include normal use, foreseeable misuse, alarms, lockout/tagout procedures, maintenance, and emergency response.
- Pilot and observe. Compare the design with real tasks, including shifts and particularly sensitive groups.
- Measure and review. Record incidents, false alarms, cancellations, breakdowns and changes, and reassess when equipment, conditions or processes change.
Practical example
In a warehouse, staff repeatedly cross an area where forklifts circulate. The company is considering installing an automated guided vehicle (AGV) and interlocking doors to separate the movement of goods and people. The measure is not validated simply because the vehicle detects obstacles. The evaluation analyzes routes, visibility, speeds, crossings, signal loss, battery life, maintenance, evacuation procedures, and predictable pedestrian behavior.
Operators and representatives are consulted, traffic flows are physically separated where possible, and an accessible stop is defined. During the pilot program, braking incidents, unplanned entry, and safety overrides are recorded. If personnel must enter the operating area to resolve congestion, a safe lockout/tagout procedure is in place. The primary indicator is not the number of journeys the system executes, but whether it reduces exposure without creating other significant risks. Automation complements traffic management; it does not justify eliminating training, signage, maintenance, or supervision.
Regulatory framework in Spain and the European Union
Law 31/1995 establishes the duty of protection, the principles of preventive action, and the integration of prevention . Articles 18, 19, and 33 require information, training, and consultation, including regarding the introduction of new technologies. Royal Decree 39/1997 mandates the reassessment of workstations affected by new equipment or technologies. Royal Decree 1215/1997 regulates the suitability, use, verification, and maintenance of work equipment, as well as training and information.
When the solution incorporates machinery, the regulations governing its marketing and commissioning also apply. As of the date of this review, Royal Decree 1644/2008 transposes Directive 2006/42/EC. Regulation (EU) 2023/1230 will be generally applicable from 20 January 2027, although some provisions have earlier dates. Product conformity and safety in its use are complementary aspects: having markings and documentation does not replace workplace assessment, consultation, or monitoring of actual operating conditions.
