Training Records vs Competence Demonstration vs Task Authorization: Which Gate Proves Readiness?

A contractor arrives for a high-risk maintenance job with a completed induction, a current training certificate, and a supervisor who says the crew is ready. The paperwork is in order. The harder question is whether anyone has seen the worker perform the critical task under the conditions that will exist on site.

Training records, competence demonstrations, and task authorization are often treated as interchangeable proof. They are not. A record shows that an activity was completed. A demonstration shows that a person can perform a defined task. Authorization confirms that this person, this task, this equipment, and these controls are acceptable at a specific time. The safest gate depends on the exposure and the decision that must be made.

Andreza Araujo's work in safety transformation supports a practical distinction. Prevention becomes credible when leaders connect what was taught to what can be observed at the point of work. That is why the question is not whether a worker has attended training. It is whether the organization has enough evidence to release the task without confusing administrative completion with operational readiness.

Evaluation criteria for choosing the readiness gate

Use five criteria before selecting the gate. First, define the consequence if the task is performed incorrectly. Second, identify whether the task is routine or changes with equipment, materials, weather, interfaces, or energy state. Third, establish whether capability can be observed directly. Fourth, decide how often readiness must be reconfirmed. Finally, name the person who has authority to stop the release when the evidence is incomplete.

These criteria separate three different management questions. Training records answer whether the organization provided instruction. Competence demonstrations answer whether the person can perform a defined activity to an agreed standard. Task authorization answers whether the work can begin now, with the current people, controls, boundaries, and operating conditions.

The three gates should work as a chain rather than a contest. A worker may need baseline training before a competence demonstration, and a successful demonstration may still be insufficient for authorization when the job includes a new interface, a temporary control, or a changed energy state. The evidence should become more specific as the decision moves closer to exposure.

Training records establish the baseline, not the release decision

Training records are useful because they create a visible baseline. They show that the organization identified a topic, assigned instruction, and retained evidence of completion. For site induction, chemical awareness, emergency arrangements, or required legal content, that baseline prevents the organization from relying on memory and informal coaching.

The problem begins when the record is used as a proxy for performance. A slide deck can explain an isolation sequence without proving that the worker can identify every energy source, apply the device correctly, verify zero energy, and respond when the equipment configuration differs from the example. The record confirms exposure to information. It does not confirm reliable execution.

ISO 45001:2018 treats competence as more than attendance because the organization must determine the necessary competence, ensure that people are competent, and retain appropriate evidence. That logic is especially important for safety-critical tasks, where a certificate can be current while the worker has never performed the practical sequence on the equipment used at the site.

Use the training record as the entry condition for the next question. If the task has serious consequences, variable interfaces, or a narrow margin for error, the record should trigger a demonstration or authorization review rather than close the gate.

Competence demonstrations test whether capability is visible

A competence demonstration converts a claim into observable evidence. The assessor defines the task, the conditions, the critical steps, the unacceptable deviations, and the standard for passing. The worker then performs the activity or a controlled simulation while the assessor records what happened.

This gate is strongest when the task contains a sequence that cannot be judged from a written test. Examples include operating mobile equipment, connecting a fall-protection system, carrying out an electrical isolation, testing a gas detector, or leading a confined-space rescue sequence. In each case, the organization needs evidence of action, not only evidence of exposure to instruction.

Competence should be assessed against the real difficulty of the work. A demonstration performed on an unloaded training rig may not prove readiness for a congested plant, a night shift, a weather-sensitive lift, or an interface with another contractor. The assessment should therefore state which conditions are covered and which changes require reassessment.

The main trap is treating a pass as a permanent property of the person. Skills decay, equipment changes, procedures drift, and workarounds become normal when production pressure rises. Reassessment triggers should include a long gap since performance, a significant incident or near miss, a process change, a new machine, a failed verification, or evidence that the worker has not performed the task recently.

Task authorization connects readiness to the job in front of you

Task authorization is the most specific gate because it asks whether the work may begin under current conditions. It does not replace training or competence. It uses them, then adds the information that belongs to the job, such as scope, location, equipment status, simultaneous operations, isolation boundaries, weather, rescue arrangements, and who has authority to stop or change the work.

A permit-to-work system is one form of task authorization, but the label matters less than the decision quality. A permit that repeats a generic method statement without checking the actual worksite is a document transfer. An authorization that confirms the people, controls, interfaces, and stop conditions is a release decision.

OSHA's Control of Hazardous Energy standard, 29 CFR 1910.147, illustrates the difference between general instruction and task-specific control. Employees must understand the purpose and function of the energy-control program, while the work itself still requires identification of energy sources, application of devices, verification, and controlled restoration. A completed course cannot substitute for the release evidence required at the equipment.

Authorization should be withdrawn when the task changes. A new contractor enters the area, a barrier is removed, another crew starts nearby, the isolation boundary shifts, weather affects the method, or the rescue plan becomes unavailable. The right response is not to edit a signature after the fact. It is to pause, reassess, and issue a new release decision.

Decision matrix: match the gate to the exposure

The matrix is intentionally conservative around high-consequence work. When the consequence is severe and the task is variable, the organization should require more than one evidence layer. This is not bureaucracy for its own sake. It reduces the chance that a person is released because one field in a learning system happens to be green.

How leaders prevent the three gates from collapsing into one

Start by mapping the tasks that can produce a serious injury or fatality, then identify which evidence is currently used to release each one. If every task is released by training status alone, the system has a visibility problem. If every task requires a permit but no one observes capability, it has an authorization problem. If demonstrations are performed but never linked to current equipment or conditions, it has a transfer problem.

Assign ownership by decision. The learning function can own curriculum quality, operations can own task standards, supervisors can verify conditions, and EHS can challenge weak evidence. The exact structure may vary, but the release authority must be explicit. A shared spreadsheet is not an owner, and a signature is not an escalation route.

Review the gates after an unexpected event, a failed control verification, or a repeated deviation. James Reason's work on latent conditions is useful here because a recurring execution failure may reveal a design or management weakness rather than an isolated individual lapse. The review should ask whether the organization trained the wrong thing, assessed the wrong behavior, or authorized work without enough current evidence.

Recommendation by context

Use training records for baseline knowledge and access decisions, but do not present them as proof that a worker can perform a safety-critical task. Use competence demonstrations when the critical evidence is observable performance. Use task authorization when the decision depends on current conditions, interfaces, controls, and a named release authority.

For high-consequence, non-routine work, combine all three. Training establishes the foundation, demonstration confirms capability, and authorization decides whether the job can begin now. That sequence gives leaders a defensible answer when someone asks the most important readiness question: what evidence shows that this person can do this task safely under these conditions?

What to check before releasing a safety-critical task

• The required training is current and matches the actual hazard and work scope.
• The worker has demonstrated the critical steps on comparable equipment or in a controlled simulation.
• The task authorization names the scope, boundaries, controls, interfaces, and stop conditions.
• The release authority can explain what evidence was reviewed and when the decision must be renewed.
• Changed conditions trigger a pause and reassessment rather than a silent edit to the original approval.