Safety-I and Safety-II complement each other
The perspective commonly called Safety-I focuses on understanding and reducing what goes wrong. Safety-II adds the study of how activities usually function. Analysing both types of situation helps reveal conditions that may go unnoticed when investigations take place only after harm.
The Royal College of Obstetricians and Gynaecologists explains that Safety-II does not replace Safety-I. Accident investigation, risk assessment and barrier verification remain necessary in OSH. Learning from normal operation does not reduce the importance of correcting known failures.
What is studied in everyday work
The focus includes coordination, available information, variations in demand and the adjustments people make. An apparently stable task may depend on conversations, checks or support that the procedure does not describe.
Observe when these resources are available and what happens when they are absent. Selecting exceptional examples of good performance is not enough: ordinary situations, including seemingly unremarkable ones, need to be understood. Distinguish what enabled safe work from what merely ended without harm because circumstances were favourable.
Useful adaptation and the normalisation of risk
People adjust their work to situations that are not always anticipated in detail. Some adjustments help maintain control; others weaken safeguards or shift risk elsewhere. Repeating a practice without accidents does not prove it is safe.
Assess its limits using technical and preventive expertise. When activity depends on extraordinary effort, memory or constant improvisation, the system may be fragile. Resilience must rest on resources, competence and room to act, rather than asking workers to absorb every pressure.
How to conduct a short review
Select a task, talk with those who perform it and observe work at different times. Ask what changes, how difficulties are detected and what support helps resolve them. Include functions that provide resources or receive the output.
- Describe variations and dependencies.
- Identify signs of loss of control.
- Check the resources and conditions needed.
- Agree changes and review their effects.
A useful review may result in strengthening coordination that works or removing a fragile dependency. It does not require turning every observation into a new procedure.
Limits of application and evaluation
EUROCONTROL has developed this approach in complex air traffic management systems. Applying it in another sector requires understanding differences in tasks, risks and organisation. Adopting the Safety-II label provides no universal guarantee of fewer accidents.
Connect findings with verifiable decisions and monitor unintended consequences. HSE distinguishes identifying lessons from turning them into effective changes. Measuring only meetings or positive stories can produce visible activity without strengthening the actual capacity to operate safely.
Hypothetical example: shift handover
A site studies several handovers that enabled changes in workload to be managed without incidents. It finds that a short conversation between supervisors prevents information being lost that the written record does not capture clearly.
The necessary information is reviewed, and time and a suitable means of communicating it are provided. Arrangements are also defined for occasions when both supervisors are not present together. The improvement preserves a useful practice and reduces dependence on particular people, without concluding that every previous handover was necessarily safe.
