Bespoke Safety Training for High-Hazard Work: Designing Competence Beyond Off-the-Shelf Courses

• A certificate proves attendance; it does not prove that someone can perform the task safely when it matters.
• Generic training rarely reflects the specific controls, pressures and failure points of your actual operation.
• In safety-critical work, training is only adequate if it matches the actual risk — attendance alone is unlikely to satisfy that test.
• Training does not reduce risk unless what was learned is applied and maintained on the job.
• If you cannot demonstrate that people perform to a required standard, you are relying on records rather than control.

The comfortable fiction is this: once a course is completed and a certificate issued, the training obligation is met. For foundational awareness — fire safety, manual handling basics and induction content — a standardised course can be a proportionate starting point.

That approach becomes inadequate when the same logic is applied to safety-critical tasks such as working at height on complex structures, operating high-hazard plant, responding to unexpected equipment or process failures and managing emergency scenarios that depend on team decision-making under pressure.

In those contexts, course completion in place of demonstrating actual competence creates a gap between visible and reliable assurance. In many cases, people believe the gap is closed. Supervisors record them as trained, and dutyholders log the hours. But none of that tells you whether, when it matters, someone can actually perform to the standard that keeps them and others safe.

That gap — between what training records show and what people can actually do — is where bespoke safety training becomes necessary rather than desirable, because the task, the specific controls in place and the specific ways it can go wrong are all particular to the operation.

The HSE’s own guidance is clear. The HSE defines competence as performing to a recognised standard on a regular basis, combining skills, knowledge and experience, and states that training contributes to competence but is not sufficient alone (HSE, n.d.).

The legal framework does not say ‘provide training.’ It says provide training that is ‘adequate’ for the task, the person and the risk, and ‘adequate’ is the word with the most practical and legal significance.

Under the Health and Safety at Work etc. Act 1974, Section 2(2)(c), employers must provide the information, instruction, training and supervision necessary to ensure, so far as is reasonably practicable, the health and safety of employees at work (HSWA 1974, s2(2)(c)).

The Management of Health and Safety at Work Regulations (MHSWR) 1999, Regulation 13 goes further: it requires employers to consider employees’ capabilities before entrusting tasks, to provide training that is adequate and appropriate, to repeat it periodically where necessary and to adapt it when risks change (MHSWR 1999, reg 13).

For those working with machinery and equipment, the Provision and Use of Work Equipment Regulations 1998 (PUWER), Regulation 9 requires that users and their supervisors receive adequate training, including awareness of the methods of use, the risks and the precautions (PUWER 1998, reg 9).

The Work at Height Regulations (WAH) 2005 make a competence requirement explicit: no one should engage in work at height unless they are competent to do so, or unless they are being supervised by a competent person. The regulation does not say ‘certified.’ It says competent (WAH 2005, reg 5).

The HSE is explicit that what counts as ‘adequate’ depends on the job, the equipment, existing competence and the level of supervision available (HSE, n.d.).

The evidence on safety training deserves careful reading — it is often cited selectively. A large-scale research review by Robson and colleagues found that training does improve occupational health and safety behaviours and practices – the evidence supports this finding.

But the same review found insufficient evidence that training alone produces large, reliable improvements in injury or illness rates. The authors were direct: significant health outcome improvements should not be expected from training based on available evidence (Robson et al., 2012).

That finding is not an argument against training — it is an argument against treating training as the primary control. Incident rates are shaped by engineering controls, supervision quality, safety climate, workload and many other variables.

Separate research by Burke and colleagues identifies a further point: not all training methods are equal. More engaging approaches involving practice, active participation and realistic scenarios consistently outperform one-way instruction, such as lectures or e-learning without interaction.

The clearest improvements are in knowledge and safe behaviour immediately after training. Reductions in injury rates — which depend on many factors beyond training — are smaller, though the research shows they are still present (Burke et al., 2006).

For high-hazard work, the goal is competence that can be observed and assessed in the specific tasks that matter — not a certificate that serves as a defence against liability. Achieving it depends on a quality the research literature calls transfer — meaning whether what is learned in training actually carries over to how people perform at work.

Transfer of training refers to how well what is learned in training carries over to job performance and is maintained over time. It is also the most consistently overlooked factor in how training is designed and delivered.

Research by Baldwin and Ford identifies three categories of factors that determine whether training carries over to the job: how training is designed and delivered (content, methods, how and how often practice opportunities are structured); factors relating to the individual learner, including their ability, motivation and confidence; and the work environment (support and reinforcement from colleagues and supervisors, and opportunities to perform the trained behaviours in the workplace).

That last factor is frequently underestimated. It is possible to design excellent training and still achieve poor transfer, if the working environment does not create regular, supported opportunities to perform correctly. For infrequent or emergency tasks, this becomes acute: someone may attend training in January and not face that scenario in practice until the following year – or at all. Knowledge fades. Memory for how to carry out a task step by step fades over time. Confidence does not reliably track actual performance.

These gaps between training and actual performance are why the HSE’s guidance on competence management places such emphasis on ongoing assurance rather than one-off training events, and why the Office of Rail and Road (ORR) explicitly cautions against relying on external qualifications as a substitute for evidence of competence within the operation (ORR, n.d.).

Competence is not something you certify once. It is something you maintain, monitor and verify.

The following framework draws on HSE guidance, ORR competence management principles and the research evidence. It is a system to be managed.

For each high-hazard role or task, begin with task and scenario analysis. Consider normal operations, abnormal conditions, conditions where equipment or systems are not functioning normally, maintenance activities and foreseeable emergency scenarios. The HSE is explicit that employers should base competence assurance on risk assessment outputs, not select training courses from whatever happens to be available (HSE, n.d.). If the course does not address the actual task in its operating conditions, it is incomplete regardless of how well the training provider has designed it.

Translate risk controls into observable performance criteria: what does good look like, what does unacceptable look like, what are the decision points and the conditions under which work must stop? Competence standards derived from procedures, risk assessments and engineering controls give assessors something concrete to evaluate against, rather than relying on the impression formed at the end of a training day.

For safety-critical tasks, one-way instruction — lectures or e-learning without interaction — is rarely sufficient. Where the consequence of error is serious, training methods should include structured practice, demonstrations of what correct and incorrect performance looks like, simulation or realistic scenarios and coached on-the-job training (OJT). The HSE explicitly supports OJT where it is structured, linked to risk assessments and controls, and supported by other methods as appropriate (HSE, n.d.).

Employers should document and standardise their competence assessment process, using qualified assessors to conduct it.

Practical demonstrations, scenario-based assessments and observation in realistic conditions are more meaningful than written tests alone — particularly where skills that go beyond technical task knowledge, such as the ability to recognise and respond to changing conditions, communication and decision-making, are relevant to safety.

ORR guidance requires standardised assessment, records that can be checked and verified by an enforcement officer or auditor, and assessor competence (ORR, n.d.).

Transfer — whether what was learned in training is applied and maintained at work — does not happen automatically. Supervisors need to create opportunities for correct performance, reinforce good practice and intervene when drift occurs.

For infrequent or emergency tasks, this may require structured, repeated practice with feedback and supervision, or structured discussion exercises that walk through emergency scenarios without enacting them, to remain current.

Employers should treat training as a risk control measure that requires active management: validated (did it deliver what it was designed to deliver?), evaluated (is this the right kind of training for our needs?) and subject to a defined refresh and reassessment cycle, especially for safety-critical or infrequent tasks.

This is explicit in HSE guidance on validation and evaluation, and mirrors the ORR’s approach to managing competence as a cycle of design, implementation, monitoring and review (ORR, n.d.).

When equipment, systems of work or task responsibilities change, employers must review and update training and competence records accordingly – a direct requirement under MHSWR Regulation 13 (MHSWR 1999, reg 13).

For contractors, the dutyholder retains responsibility for verifying that competence arrangements are in place and adequate. HSE guidance on contractor management is clear that the dutyholder must retain ‘intelligent customer capability’ – that is, the ability to assess and govern what has been contracted out (HSE, n.d.).

Off-the-shelf completion as proof of competence for safety-critical tasks is not supported by UK guidance, UK law or the research evidence. HSE guidance consistently treats training as necessary but not sufficient and expects stronger assurance for higher-hazard work. Training that relies solely on one-way instruction, such as e-learning or classroom presentations without practice, is similarly insufficient where practice and realistic assessment are central to competence development.

Many organisations persist for understandable reasons: cost pressure, time pressure, the availability of ready-made solutions and the genuine difficulty of building bespoke systems from scratch. Training with a completion certificate generates records — and it is easier to maintain the appearance of assurance than to build a system that actually provides it.

Training records tell you that people attended. Competence evidence tells you they can perform. In high-hazard work, only the second one actually reduces risk and the systems that generate it look quite different from a course catalogue.

Bespoke safety training is not about custom branding or rewriting slides. It is about designing competence around a specific task, specific safeguards and specific failure modes. High-hazard competence cannot be generic because the risks are not generic. The moment you define the task and its credible degraded scenarios properly, the limitations of off-the-shelf coverage become obvious.

If there is one practical step worth taking first, it is this: pick your highest-consequence task and ask, honestly, what evidence you currently have that the people doing it can perform to the required standard under realistic conditions – not just that they once attended something relevant. The gap between what you find and what you need is where the design work begins.