Safety inspection:
surprise or scheduled?

Access routes are cleared. Missing guardrails appear in place. Documents are arranged on the table, and the right people are waiting to give the right answers.

But what happens the following morning, when no one is expected to visit?

An occupational safety inspection can measure workplace safety. It can also measure how well an organisation can prepare a convincing performance for an inspection.

Should an inspection therefore be announced in advance, or should the inspector arrive without warning?

The answer is not simple. Scheduled and surprise inspections serve different purposes. An effective inspection programme generally needs both.

Here I am referring especially to site and safety inspections carried out by the organisation itself, a client or a safety specialist. Inspections conducted by regulatory authorities follow their own procedures and statutory powers.

A scheduled inspection supports cooperation

When safety is managed well and cooperation works, a scheduled inspection should generally be the primary method.

An agreed time allows the right leaders, supervisors and specialists to be present. Relevant documents can be made available, and participants can prepare questions they want to discuss.

The inspector then has a better opportunity to understand how the organisation’s safety management system is designed and how it is intended to work.

An inspection should not feel like a raid. At its best, it is a joint opportunity for the organisation and the inspector to examine work, identify weaknesses in the system and find practical improvements.

The inspector is not an enemy arriving to find someone to blame. They should be an independent but constructive partner whose role is to help make work safer.

This approach depends on trust. Trust cannot mean that the effectiveness of agreed practices is never verified.

A surprise inspection reveals an ordinary working day

A scheduled inspection is well suited to showing how the safety system has been designed. It allows procedures, maintenance records, risk assessments and the progress of corrective actions to be reviewed.

At the same time, it inevitably measures the organisation’s ability to prepare for an inspection.

A surprise inspection has a different purpose. It shows whether agreed controls remain in use when no one is preparing for an external visit.

Are the same methods used in practice as in procedures and induction materials? Are access routes normally clear? Is fall protection used? Are guards in place? Do work planning and supervision function even when an inspector is not present?

When the inspector was restricted after an inspection

I once carried out an unannounced safety inspection at a bridge construction site. The site was untidy and disorganised. Guardrails were missing, and working platforms had dangerous openings. These were not minor defects but visible physical hazards capable of causing a serious accident.

Obtaining information was also difficult. I could not get an adequate picture of the situation, and cooperation was limited. I had to prepare my report using relatively restricted information. There was also an impression that more serious problems may have been kept out of sight.

One might have expected management to focus on the missing guardrails, dangerous openings and weak cooperation. That did not happen.

Management supported the site manager. After the inspection, I was told that I could no longer visit worksites without giving at least two days’ notice.

Advance notice is not inherently a problem. Worksites have legitimate access-control, induction and safety requirements. Scheduled inspections are also often the most effective way to develop operations.

The decisive issue was the context of the decision.

The restriction was imposed immediately after serious shortcomings and unwillingness to cooperate had been identified. Management’s strongest response was directed not at the hazards, but at the inspector’s ability to see them without warning.

The decision did not install a single guardrail, close an opening or improve cooperation. It only gave the site more time to prepare for the next inspection.

The inspection had therefore revealed more than unsafe conditions. It showed how the organisation responded to uncomfortable information.

When a poor score threatens a reward

I encountered another example on a much larger bridge project where I carried out an MVR measurement.

MVR is a systematic observation method used on Finnish civil-engineering worksites. Observations are recorded as correct or incorrect, producing a score that represents the site’s safety level.

The measurement itself was routine, but the result was poor. It was initially accepted at the site. Later, however, I was informed that I was no longer allowed to conduct MVR measurements. The poor result affected performance bonuses.

The purpose of the measurement was to identify shortcomings and support safety improvement. Once the result threatened financial rewards, the site’s safety level was no longer treated as the problem. The measurement and the person conducting it became the problem.

The worksite did not become safer by preventing further measurements. Only the visibility of the problem decreased.

This is a significant risk when safety indicators are linked to rewards, reputation or career development. People may gradually move from managing risks to managing numbers.

Exceptional conditions can be arranged for an inspection. A poor score can be challenged even when good scores are accepted without discussion. The inspector’s movement or access to employees can be restricted. The person who produced an uncomfortable result can be replaced, or the measurement can be stopped altogether.

Connecting safety results to incentives is not automatically wrong. The problem arises when those who benefit from the reward can also control the measurement, the selection of the assessor or the handling of the result.

People whose reward depends on a safety measurement must not be able to decide alone who conducts it or how its results are handled.

Otherwise, the system creates an incentive to suppress bad news. The number may improve while the underlying risk remains unchanged.

The inspector’s independence is not an attack on the organisation. It also protects management from making decisions based on information that is overly positive or deliberately polished.

Warning signs include selectively challenging poor results, unjustifiably restricting the inspector’s movement, controlling who may be interviewed or replacing the inspector who produced an uncomfortable result.

A single sign does not prove deliberate concealment. It does, however, provide a valid reason to verify whether the conditions seen during an inspection represent an ordinary working day.

A surprise inspection is not covert surveillance

An unannounced inspection can also be carried out badly.

The surprise should concern the timing of the inspection, not the inspector’s conduct. The visit need not be announced days in advance, but on arrival the inspector must follow access control, induction and other safety requirements.

A surprise inspection must not become covert surveillance.

I once knew a safety manager whose credibility suffered after he was seen hiding in bushes with a camera, waiting for employees to make mistakes.

The image is amusing, but it also demonstrates exactly how an inspection should not be conducted.

An inspector should not hide behind a fence or in bushes to photograph workers in secret. The timing may be unexpected, but the inspector’s conduct must be open.

They should introduce themselves, explain the purpose of the visit and seek to establish a professional and constructive dialogue.

Constructive does not mean uncritical

When a shortcoming is identified, the first response should not automatically be an accusation.

The inspector can begin by asking: Is this the normal way the task is performed? What made the safer method difficult to use? Was the necessary equipment available? Did the schedule, resources or supervision influence the situation?

People do not usually come to work intending to act unsafely. They adapt to the tools, resources, instructions, time pressures and expectations around them.

Understanding these conditions does not mean accepting unsafe work. If an immediate danger is found, it must be controlled first. A constructive approach does not mean allowing a serious hazard to continue while it is discussed.

Once the danger has been controlled, the organisation must establish why the situation arose and how recurrence will be prevented.

The purpose of an inspection is not merely to catch an individual employee making a mistake. It should help the organisation understand work and improve the system.

A friendly and professional atmosphere usually produces more information than fear. Better information enables better decisions.

Friendliness is not weakness, and a constructive approach is not uncritical. An inspector can treat people with respect while preserving independence.

Serious shortcomings must be reported clearly, repeated neglect must be addressed, and management must accept responsibility for conditions it allows to continue.

An effective inspection programme needs both methods

Scheduled inspections should generally form the foundation of an inspection programme when cooperation works and safety is managed well. They can build trust, examine the management system thoroughly and support joint improvement.

Surprise inspections are needed when cooperation is weak, the safety culture is still immature or the organisation appears to manage the impression created during an inspection rather than safety itself.

They can also be used as occasional spot checks in well-functioning organisations. The purpose is not to assume that everyone is dishonest. Every management system nevertheless needs an independent way to verify that planned practices and workplace reality correspond.

The assessment criteria must always be clear, even when the exact timing is not announced. An inspection must not be a trap in which employees are judged against unknown rules.

A poor result must not lead to the person who produced it being removed from the task.