Calibration and Control of Measuring and Monitoring Devices

FSQMS Guide

Calibration and measuring-device control establish whether the results used for food safety, legality and quality decisions are reliable for their intended purpose. Calibration provides evidence about the relationship between an instrument's indication and reference values. The wider control programme identifies equipment, defines acceptable performance, checks its condition and response, prevents unauthorised adjustment and addresses measurements already made when reliability is lost.

An instrument can have a current certificate and still be unsuitable for a particular decision. Its accepted range may exclude the value being measured, a replacement probe may change the measurement chain, or the method may produce too much uncertainty to demonstrate achievement of a critical limit. Effective control connects each device and configuration with its actual use, current evidence and the people authorised to interpret and act on the result.

Establishing the measurement-control programme

Begin with the measurements on which the site relies. Follow critical control point (CCP) monitoring, product-safety checks, legal conformity decisions and quality assessments to the equipment producing each result. Include installed sensors and recording systems, portable instruments, laboratory equipment, working standards and references. Identify spare, borrowed or temporary equipment when its results will support the same decisions.

For each use, establish the quantity measured, units, working range, conditions and decision supported. Obtain the relevant process or product requirement and determine how accurately and consistently it must be measured. A broad quality tolerance and a narrow safety decision can require different capability even where both measurements use a temperature instrument. Select equipment and methods against those needs before relying on results.

Requirements, responsibilities and usable resources

When interpreting measurement requirements, identify the legislation, customer commitments and adopted assurance conditions that apply to the equipment, products, activities and markets concerned. Check the actual requirement, current edition, effective date and scope. Translate it into requirements for measurement performance, methods, checking and evidence. Obtain competent interpretation of unresolved applicability or conflicting conditions; an internal approval cannot waive a binding duty.

The following functional allocation distinguishes the responsibilities needed for measurement control. The technical or quality function defines and maintains measurement-control requirements, approves criteria and methods and reviews consequential evidence. Engineering or the assigned asset custodian maintains physical identity and location, coordinates calibration and performs authorised technical work. Competent checking personnel perform routine comparisons. Users maintain protection and recognise unreliable status. Operations implements approved measurement arrangements and product containment. These functions can be combined in one person or distributed across teams, provided their decisions and limits remain clear.

Assign who can change measurement-affecting settings, accept calibration evidence, determine return to use, approve an alternative monitoring arrangement and decide affected-product disposition. A provider's calibration service does not transfer the site's product responsibility. Senior management is responsible for providing the resources and resolving competing priorities, subject to the technical requirements and applicable legal conditions.

Assess actual capacity for checking, stabilisation, recording, review and failure response. Allow for production access, provider lead time, reference availability, transport and reinstatement. Spare equipment can maintain continuity only if its performance and current status are suitable for the measurement and an approved method and competent users are available. A paper register or controlled spreadsheet can support the programme where identity, updates and retrieval are reliable; specialist software is an implementation choice.

Train all personnel who approve, review, perform, maintain, use, assess, investigate or escalate this work to the level their duties require. Provide instruction on appointment, following significant changes and after deficiencies revealed by checking, calibration review, audit, failure, investigation or incident. Observe practical comparisons and exception decisions to establish competence. Attendance alone does not demonstrate the ability to recognise a failed result or use a reference correctly. Deputies need current instructions, evidence access, competence and authority within defined limits.

Device identification, status and protection

Equipment identification connects the physical device with its approved measurement purpose, current location and performance evidence. Users need to be able to establish whether a particular device is currently suitable and authorised for the measurement they intend to make. Similar appearance or a familiar position on the line cannot establish that relationship.

Registering devices and measurement chains

Maintain a documented list of in-scope equipment and its location. Give each device a persistent identification code applied to, attached to or otherwise reliably associated with it. Link that identity to its approved uses, calibration due date, checking requirements and current status. The asset register and calibration schedule can share one system while retaining their distinct purposes and update responsibilities.

For connected equipment, identify the combination on which the result depends. A sensor, transmitter, indicator, recorder and software conversion can each affect the final value. State whether the accepted calibration covers the complete chain or only a component. A calibrated probe connected to the wrong scaling or recording channel can produce a misleading result. Record the relevant configuration so that replacing a component prompts assessment of its effect on the measurement chain.

Enter new and replacement equipment into asset and calibration controls before first relying on its results. Equipment acceptance addresses installation and readiness as well as measurement capability. A replacement that meets preapproved requirements for the same use can follow the established acceptance route. Engineering updates its identity and asset information and promptly notifies the person responsible for the calibration programme. An unassessed substitute cannot inherit another device's certificate or status.

When equipment moves, update its recorded location and check whether the new environment or installation changes measurement performance. Notify the person responsible for measurement-control requirements before relocation or removal leaves a required measurement without approved equipment. Reassignment to a different range, process or decision needs technical review of suitability, criteria and checking requirements before relying on the results.

Making current status usable

Distinguish equipment accepted for defined use from equipment awaiting assessment, overdue, failed, under repair, restricted or retired. Use status information the operator can interpret at the point of use. Where equipment is calibrated daily or more frequently under the approved method, current instructions and completed calibration records can demonstrate status without a due-date marking. The device still needs an unambiguous identity and retrievable current evidence.

Before using a result, match the physical device to its recorded identity, approved use and current status. Check its relevant protective controls and whether the indication, units and scale are readable. Resolve a missing identity, overdue or unverifiable calibration, damaged display or unexplained setting change with the person authorised to assess measurement suitability. Stop use where reliability cannot be demonstrated. Printing another label cannot resolve the underlying evidence gap.

A restriction needs a practical boundary. An instrument accepted only for a particular range or quality use should have supporting evidence and an instruction that prevents use outside that scope. A limited-use label that does not identify the permitted application leaves the operator unable to determine whether the intended measurement is allowed. On retirement, withdraw use permission and prevent unintended return while retaining the identity and history needed to interpret previous measurements.

Preventing adjustment, damage and deterioration

Restrict measurement-affecting settings to authorised technical personnel or the authorised provider. Suitable controls can include permissions, locked menus, seals or physical access restrictions. Compare actual access with the assigned rights and check that permitted work remains feasible. A shared administrator password or an unchecked seal can defeat the intended restriction. Information-system access alone does not confer authority to alter a measurement.

Protect equipment during use, movement, cleaning and storage against the exposures it actually faces. A balance needs stable support and suitable loading; a probe needs handling and cleaning compatible with its design. Examine moisture, impact, damaged leads, contamination and unsuitable storage where they could change performance. Keep protective instructions usable during normal production rather than relying on ideal handling that the workplace cannot support.

An impact, broken seal or exposure outside approved conditions can call reliability into question before the next scheduled check or calibration. Stop relying on the device and report what happened and when to the person authorised to assess measurement suitability. That assessor determines whether a comparison or engineering assessment can establish current suitability and whether previous measurements need review. Follow the out-of-limit response where reliability is lost or uncertain.

Checking, calibration and adjustment

The checking programme establishes whether identified devices, including new equipment, meet the performance requirements for their measurements. Calibration characterises the relationship between reference values and indications, including associated uncertainty. Adjustment changes the instrument's response. A conformity check applies specified criteria to the evidence. Keep these purposes distinct so a successful adjustment does not conceal an adverse original result.

Defining a reproducible comparison

The person responsible for technical measurement requirements defines the quantity, units, relevant range and comparison points, acceptable error and any repeatability requirement. Select comparison points around values that determine important decisions and across the range actually relied upon, with sufficient coverage to reveal relevant response changes. A single point may provide a defined limited check; it cannot establish performance throughout an untested range.

Specify the reference, preparation, comparison conditions, stabilisation, readings and calculation. For temperature comparisons, the reference and tested sensor need sufficiently comparable conditions, with immersion and stabilisation appropriate to their designs. For balances, define suitable comparison loads, positioning and stable support. Apparent disagreement can arise from a poor test as well as an unreliable instrument. The method needs enough detail for competent people to obtain and interpret comparable results.

Calculate indicated error as instrument indication minus reference value, using the reference value corrected as its certificate requires. For example, an indication of 80.6 °C against a reference value of 80.0 °C gives an indicated error of +0.6 °C. If the approved method permits an error magnitude no greater than 0.5 °C under those conditions, the result fails. The additive correction at that point would be −0.6 °C; applying it to interpret a result differs from physically adjusting the device. Any correction used operationally needs an approved basis covering the intended measurement. One point's correction cannot automatically be extended to the entire range.

Repeatability concerns variation under defined repeated conditions. A device consistently reading +0.6 °C high can be repeatable and still unsuitable. A small average error can also conceal excessive variation. Where repeatability matters, define the repetitions and criterion and assess actual results accordingly. Do not average away a failed reading unless that statistic is the basis of the approved method. Resolution describes the smallest distinguishable indication, not demonstrated accuracy.

Explain how measurement uncertainty affects the decision, particularly for borderline results. Predetermine how the method treats uncertainty and the acceptance boundary. Rounding or repeating a test until one reading passes does not establish conformity. When a test is suspect, preserve its result, investigate the reference and conditions and repeat under corrected conditions with the reason recorded.

Use a defined method traceable to recognised national or international standards where possible. If such linkage is unavailable for a particular checking or adjustment method, the technical approver should document an alternative capable of demonstrating the required accuracy and relevant repeatability for the intended use. Convenience is not performance evidence. Apply any stricter legal or contractual condition. This alternative does not remove the requirements under Reference equipment and traceability.

Setting and delivering risk-based intervals

Predetermine checking and necessary adjustment frequencies through documented risk assessment. Consider the consequence of an incorrect result, intensity and conditions of use, drift and failure history and relevant manufacturer or calibration advice. Include new equipment and event-driven checks after conditions that could invalidate status. A contractor's annual visit does not by itself justify the interval.

Work backwards from calibration expiry to allow booking, transport or access, service, evidence review and reinstatement. The person responsible for the calibration programme monitors forthcoming dates and alerts engineering or the calibration coordinator early enough to complete this work. Where a required check or calibration is overdue or unverifiable, remove the device from the affected use until acceptable evidence establishes status. A later satisfactory comparison does not automatically prove the intervening measurements were reliable.

Review intervals against actual as-found performance and use. Shorter intervals or a changed method can be necessary where errors approach the limit or damaging conditions recur. Any decision to lengthen an interval needs technical justification before the revised interval takes effect and must remain consistent with applicable obligations. Keep the earlier history and original due dates; an extension cannot retrospectively legitimise missed work.

Recording results and restoring performance

Before a comparison, obtain the current method, identify the device and reference and confirm that the reference has a valid calibration status for the comparison and that the required test conditions are met. Record the actual as-found values at specified points and compare them with the criteria. A satisfactory routine check can support continued use under the approved method without an additional management approval for every result.

Keep the device unavailable after a failed or uncertain result while the authorised technical assessor determines the implications. Preserve the as-found condition before authorised adjustment or repair. Engineering manages the technical work, internally or through an approved provider, and records the change. Obtain acceptable post-work checking or calibration over the affected scope, including points or channels the intervention could have changed.

After failure, adverse calibration, adjustment, repair or technical assessment, obtain confirmation of suitability from the authorised technical reviewer before returning the device to use. Compare the evidence with the intended measurement, not merely the provider's generic pass statement. If a device cannot meet the approved criteria, keep it out of use for that measurement or restrict its use on a technically justified basis that can be enforced. The maintenance process supports restoration while measurement acceptance retains its own decision.

Commissioning and reviewing calibration services

Specify the quantity, range, points, performance needs, uncertainty information and relevant as-found and as-left evidence when commissioning service. Assess the provider's competence and any applicable accreditation requirement for that work. A general supplier approval does not establish that a particular calibration covers the required scope.

On receipt of the calibration report, match its device identity and service scope to the device submitted and the service requested. Examine the result, reference-standard linkage, corrections, uncertainty, limitations and available findings before and after adjustment. Obtain clarification of missing comparison points or report qualifications that could affect suitability. Record the review conclusion and any restriction, then update the due date and equipment status where applicable. An acceptable as-left result supports renewed use; an adverse as-found result still requires assessment of previous measurements.

Reference equipment and traceability

Reference equipment provides values against which working devices are checked or calibrated. Its performance can affect many downstream decisions. Identify, calibrate and protect it and retain the links needed to establish both its own suitability and the consequences of a later adverse result.

Establishing and maintaining the reference relationship

Calibrate reference measuring equipment with traceability to a recognised national or international measurement standard and retain the records. Identify the reference in asset and calibration controls, including its calibration due date, current status, approved methods and scope of use. The technical reviewer checks the calibration evidence before the reference is used. Engineering or the calibration coordinator arranges recalibration before the due date.

Metrological traceability concerns a measurement result and its documented chain of calibrations to a stated reference, with uncertainty contributions. Calling an instrument traceable is practical shorthand for using it within that supported relationship. A supplier name, sticker or general statement does not establish the actual result's basis. Keep the assigned values, uncertainty, conditions and evidence needed to support the comparisons performed.

Traceability alone does not establish fitness for purpose. Compare the reference's range, uncertainty and conditions with the working-device tolerance and method. A documented chain with excessive uncertainty can leave the required conformity decision unresolved. Seek a more suitable reference, method or competent interpretation where necessary; an arbitrary reference-to-device accuracy ratio is not a substitute for the assessment.

Store, handle and use references so their assigned performance remains relevant. Restrict use to the approved comparison method and scope. Plan shared-reference availability, transport and calibration downtime against the checks due across departments. A suitable alternative reference or service can provide continuity where its evidence supports the same measurement needs.

Tracing the effect of an adverse reference result

Record the reference actually used for each working-device check. Remove a reference from use when calibration is overdue, adverse or unverifiable, traceability to a recognised standard cannot be confirmed or the reference's condition calls its reliability into question. Restoration of that reference addresses future use; it does not establish the validity of earlier dependent checks.

The authorised technical assessor identifies devices checked since the last satisfactory relevant calibration of the reference. Retrieve their results and examine whether the reference's error and limitations could change the conclusions. Keep devices unavailable where their suitability cannot be demonstrated, and arrange fresh checking or calibration using acceptable references. Coordinate the assessment across every team relying on the reference, including shared or portable equipment.

Where earlier working-device measurements may have been unreliable, assess affected products through the failure and product-control process. Maintain an owner for each unresolved device and product assessment. Completing reference recalibration does not close those downstream decisions automatically.

Measurement suitability at critical limits

Where a device assesses a critical limit, consider calibration uncertainty in deciding whether it can reliably demonstrate that the approved limit has been achieved. Obtain the approved limit from the Hazard Analysis and Critical Control Point (HACCP) system and the actual monitoring instruction. Keep the limit, monitoring design and instrument suitability as distinct but connected decisions.

Assessing uncertainty in the actual monitoring method

Measurement uncertainty describes the dispersion of values that can reasonably be attributed to the measured quantity from the available information. It is not the same as a known signed error or an absolute guarantee of a maximum possible error. Calibration uncertainty concerns the calibration relationship. Consider applicable corrections and remaining uncertainty without counting the same contribution twice. Actual use can add effects from resolution, repeatability, drift, installation and the conditions of measurement.

An accurate sensor can still give an unsuitable process result if it is too slow for the changing condition or measures in the wrong location. Examine the complete monitoring arrangement and compare it with the assumptions of the calibration evidence. Do not present certificate uncertainty alone as total uncertainty of actual use without considering additional contributions. Where assessing these effects on the measurement decision exceeds available competence, obtain metrology support for the specific quantity, range, limit and operating conditions.

The authorised technical assessor establishes a justified acceptance rule consistent with the requirement and the validated monitoring arrangement. An illustrative conservative rule for a lower limit L uses a corrected result y and a justified expanded uncertainty U, at its stated coverage, and accepts achievement only when y − U is at least L. If the illustrative limit is 75.0 °C, the corrected result is 75.4 °C and the assessment gives U = 0.6 °C, the lower end is 74.8 °C. Achievement is not demonstrated under that rule. For an upper limit, the analogous adverse comparison uses y + U against the maximum permitted value.

These numbers explain a decision calculation. They are not a universal process temperature, prescribed uncertainty, statistical guarantee or mandatory decision rule. The site's actual method needs an evidence-based rule and suitable uncertainty assessment. Record the calibration uncertainty considered, relevant use effects, rule, scope, conclusion and limitations in the calibration review or associated HACCP evidence.

Resolving unsuitable or changed measurement capability

If suitability cannot be demonstrated, first assess alternative equipment meeting the approved measurement requirements. Operations confirms that competent personnel can perform the alternative method at the required rate and location. A technically accurate substitute is insufficient if it cannot deliver the required monitoring result in time.

If suitable equipment cannot resolve the measurement-suitability question, refer the monitoring arrangement to the food safety team with the limit, method, evidence and unresolved issue. Any change to a critical limit or validated monitoring arrangement needs technical assessment, validation and approval through the HACCP system. An instrument limitation cannot itself justify relaxing the limit. The critical-limit guidance explains that separate control decision.

Reassess suitability when the device, configuration, uncertainty, critical limit, installation or validated monitoring changes. Coordinate the approved equipment, method, instructions and competence before the changed arrangement takes effect. Retain the earlier decision for its original scope. While the question remains unresolved, restrict the dependent work or use an authorised alternative that can demonstrate the approved requirement.

Out-of-limit response and affected product

A failure response addresses both equipment condition and the acceptability of measurements already made. Maintain a documented method specifying immediate actions, assessment, containment, escalation and return-to-use requirements. Restoring a device establishes a basis for future measurement; affected products need a separate supported judgement.

Stopping reliance and establishing the affected period

On detecting an out-of-limit result, damage, unreadability, unauthorised adjustment or uncertain calibration status, stop using the device and promptly report the finding to the authorised technical assessor. Identify it as unavailable and prevent unintended reuse. Preserve actual results and relevant conditions before adjustment, including failure findings from the checker or engineering provider.

The assessor identifies the last relevant satisfactory equipment check and establishes the period during which results may have been unreliable. Examine check and calibration history, incidents, configuration changes and production records. A later measured error does not prove the same error existed throughout the period. Use evidence to establish any narrower boundary; where timing remains uncertain, retain a conservative assessment of the period and products potentially affected.

For example, if the last relevant satisfactory check was at 10:00 and failure is found at 12:00, initially assess the output relying on the device during that interval and any continued affected work. Include work in progress, rework, stock movements and subsequent production wherever decisions about those products relied on the device. Continue the assessment to cover production until reliable monitoring resumes. A repair completed at 12:30 cannot establish the acceptability of the earlier batches.

Assessing safety, legality and product status

Operations identifies potentially affected product, places it on hold and prevents unintended use, movement or release pending assessment. If the affected product cannot be identified with certainty, hold all product that may have been monitored with the unreliable device. Use production and traceability records to connect times, equipment and lines with batches, work in progress and dispatches.

Assess food safety and legality separately, using the decision the measurement supported, possible error direction and magnitude, variation and relevant uncertainty. A high-reading thermometer can falsely support a minimum temperature, while a low-reading instrument can falsely support a maximum temperature. Independent measurements or process evidence can inform the judgement only where their scope and reliability address the affected condition. A later product test cannot automatically reconstruct an earlier process measurement.

Where product safety or legality depends on equipment found to be inaccurate, action must prevent at-risk product being offered for sale. Quality-only failures still need assessment against specifications and customer commitments. The person authorised to decide product disposition determines the permitted outcome from sufficient evidence, including any justified conditions for rechecking, reworking, rejecting or releasing product. Use non-conforming-product control for identification, segregation and authorised disposition while retaining the equipment failure evidence.

Where unreliable equipment supported a critical control point, treat the condition through the CCP deviation response. Determine the affected period from the last satisfactory relevant monitoring result and equipment evidence, hold all potentially affected product and restore a suitable measurement arrangement before monitoring resumes. Where potentially affected product has moved beyond direct site control, promptly initiate incident assessment to decide withdrawal, recall or neither. Begin the relevant CCP deviation or incident response without waiting for repair, the next contractor visit or a periodic management review.

Restoring equipment and resuming work

Engineering coordinates adjustment, repair or replacement through approved methods and maintenance arrangements. Acceptable post-work checking or calibration is needed over the affected scope. Obtain confirmation from the authorised technical reviewer that the approved measurement criteria are met before returning the device to use. Update the device identity, calibration due date, status and limitations as applicable, and communicate the decision to the people performing monitoring.

Operations separately confirms readiness to carry out the approved monitoring, including equipment availability, method, competent personnel and capacity. Product on hold retains its status until the authorised disposition decision. A satisfactory as-left certificate or permission to restart measurement does not release earlier product.

For a temporary alternative, define its scope, the method and its required capability, competent users, operating limits, records, the person responsible and the condition for ending or reviewing it. Authorise it only where it can demonstrate the approved requirement and meet applicable duties. Check its operation while the permanent remedy proceeds. On restoration, withdraw temporary instructions no longer applicable, communicate the current method and reconcile the temporary records and decisions with normal records.

Escalating and completing the response

Escalate significant product risk, uncertainty about the affected products or period, unavailable expertise or authority, recurring failure or delay threatening control. Identify the device and configuration, actual findings, last reliable evidence, products and period affected, protective actions in place and remaining uncertainty. State the decision or resource required and the deadline for a response. Confirm that someone with the authority to act has received the escalation. Keep an identified person responsible for the unresolved work until the recipient accepts that responsibility. Use an authorised deputy or escalate to a higher level of authority when the normal recipient is unavailable.

Determine whether significance or recurrence requires formal corrective action. Examine the evidence for causes in equipment design, protection, method, competence, use, scheduling and resources before selecting a response. Repeated damage may need changed handling or storage; repeated lateness may need capacity or provider changes. Further training or another calibration does not necessarily address those causes.

Complete outstanding work on equipment, product, incidents, instructions, training and effectiveness, or explicitly transfer responsibility for that work. The calibration case can then close while a linked action to prevent recurrence awaits later operating evidence. Retain the original failure, unsuccessful interventions and reasons for decisions. New contrary evidence or recurrence requires renewed assessment with that history preserved.

Records, verification and continuing improvement

Measurement records allow a reviewer to reconstruct the result, equipment status and decision. Maintain genuine, timely evidence and evaluate both individual execution and the continuing programme. A schedule establishes planned work; marking a task complete does not by itself establish that the device was suitable in actual operation.

Preserving and retrieving decision evidence

Identify the device and reference, method and version, date, person performing the comparison, relevant conditions, comparison points, reference values, actual indications, criteria and conclusion in the comparison record. Include repetitions and calculations needed to interpret it. Preserve as-found results, authorised adjustments or repair and as-left confirmation. Record the suitability decision, restrictions, updated status and evidence considered during the review. In the critical-limit assessment, connect the approved limit and monitoring arrangement with the corrections, uncertainty and decision rule.

Record observations as work occurs. Preserve the original entry, author, timing and reason when correcting a record. Identify later reconstruction and its evidence rather than presenting it as a contemporaneous measurement. A new form version changes future recording; it does not retrospectively change earlier results.

Retain calibration, reference and product-assessment evidence under applicable legal, customer, assurance and operational requirements, considering product life, investigation and continuing equipment history. Protect against loss or unauthorised amendment and maintain readable retrieval after provider or system changes. Follow a measurement decision back through its device, reference and comparison evidence and forward to any affected-product conclusion. The record-control guidance develops those supporting controls.

Testing capability and actual implementation

Assess whether the proposed comparison method can distinguish acceptable from unacceptable performance for its quantity, range, conditions and criteria. Use established suitable methods where they answer the need; obtain further technical evaluation where a material capability question remains. This assessment differs from process validation for the food hazard and from verification that a later comparison followed the method.

To examine programme coverage, trace measurements actually used for decisions to the equipment list, approved use, current status and evidence. Then select listed devices and locate them physically. These two directions reveal omitted instruments and obsolete entries. Include relevant portable, spare, installed and shared equipment, shifts and operating conditions. Findings from a targeted high-risk sample identify those weaknesses; they do not establish their statistical prevalence across all equipment.

Observe handling, setting access and the user's ability to establish status and respond to uncertainty. Follow a completed comparison through reference suitability, actual readings, criteria, conclusion and status update. For an adverse result, follow its product assessment and the separate return-to-use and disposition decisions. Retrieve a reference's dependent checks to establish whether an adverse reference result could be traced through the actual records. Clearly identify any exercise as an exercise rather than creating a false operational failure record.

Reviewing performance and verifying improvement

Review as-found error, drift, variation, missed checks, repeated repairs and service delays across comparable uses. Retain adverse results and distinguish devices adjusted before their original performance was recorded. A pass rate can appear favourable when the evidence conceals repeated restoration or excludes missed work.

If a rate is useful, divide failed scheduled comparisons by completed comparable comparisons for the stated devices and period and multiply by 100. Three failures among 120 completed comparisons is 2.5%. Report missed checks separately, define whether event-driven tests are included and retain information about failure severity and product consequences. Explain changes in device mix, criteria or data collection before interpreting the trend. The rate alone cannot establish a site-wide statistical failure probability.

Use successful practice, clearer information, better equipment protection and improved service coordination to strengthen capability alongside corrective work. Define the intended improvement and how it will be assessed, implement the justified change, issue updated instructions and establish competence. Verify implementation separately from the later effect under representative use. A revised method and training record demonstrate delivery; observations and later performance establish adoption and effect. Leave the effectiveness judgement open until evidence is available from the relevant operating conditions.

Report significant unresolved risk, recurrent constraints and resource needs to management in time for decisions. Technical and operational teams retain responsibility for live control. Reassess methods, intervals, references and responsibilities as equipment, products, conditions and knowledge change, preserving the evidence on which earlier decisions were made.

Stay connected with Food Industry Hub

Receive Food Industry Hub updates, resources and practical content for food manufacturing businesses and professionals.

Stay informed about new resources, articles, tools and Food Industry Hub service developments.

Raise the standard for how food gets made.