Validation

Here is a technically robust and professionally structured Validation Document for the use of thermal imagery as a valid method for temperature monitoring and recordkeeping at Critical Control Points (CCPs) within a Food Safety Management System (FSMS). This document includes methodology, calibration standards, sample validation design, and FSMS-aligned justification for its adoption and reliance.

---

Food Safety Management System (FSMS) – Validation Documentation

---

1. Purpose

This document outlines the validation methodology and evidence base for the use of infrared thermal imaging as a primary method for monitoring and recording temperature data at designated Critical Control Points (CCPs) in accordance with HACCP principles, ISO 22000, and regulatory expectations.

Given the high level of reliance on temperature data generated by thermal imaging for compliance at CCPs such as cooking, hot holding, cold holding, time control, and incoming goods, this validation aims to:

• Demonstrate the accuracy and repeatability of thermal temperature readings

• Establish confidence in the data used for decision-making and audit evidence

• Confirm that thermal imaging is a fit-for-purpose monitoring method under the FSMS

---

2. Scope

This validation applies to:

• All thermal cameras and sensors used within the facility or mobile inspection units

• Image-based temperature capture workflows processed via the Didge Platform

• CCPs relying on thermal data including:

  • CCP #1: Incoming Goods

  • CCP #3: Cooking (surface temperature of wet dishes)

  • CCP #4: Time Control (initial temperature verification)

  • CCP #5: Hot Holding

  • CCP #6: Cold Holding

---

3. Validation Objectives

• Ensure that the temperature readings from thermal imagery fall within ±1.0°C of a NATA-traceable reference device

• Validate AI-supported field population and timestamp integrity in the Didge Platform

• Confirm that the image-to-data pipeline is repeatable and suitable for audit recordkeeping

• Validate against a real-world operational environment (not lab-only conditions)

---

4. Calibration Standard

4.1 Calibration Device Requirements

All thermal cameras used must be:

• Calibrated against a NATA-certified thermal black body or equivalent traceable reference device

• Subject to three-point calibration at:

  • 0°C (chilled)

  • 50°C (intermediate)

  • 100°C (high-temperature end)

• Maintained with calibration certificates and recalibrated:

  • Every 6 months, or

  • After any impact or device update

4.2 Calibration Tolerance

• Permitted variance: ±1.0°C across all validation points

---

5. Validation Methodology

5.1 Sample Size and Test Design

• Minimum test sample:

  • 30 samples per CCP type, with a mix of foods and surface conditions

• Test conditions must represent:

  • Actual service trays, cooking vessels, delivery boxes, hot boxes, cold wells, etc.

5.2 Validation Procedure

For each CCP:

1. Capture a thermal image using the operational camera

2. Immediately capture a reference temperature at the same point using a calibrated NATA-traceable thermometer

3. Record:

   • Thermal reading from image

   • Reference temperature

   • Image file and submission timestamp

   • AI-populated data from the Didge Platform

4. Repeat across:

   • Different ambient environments

   • Multiple operators

   • Varied food types (wet, dense, light, complex surfaces)

5.3 Analysis

• Compare all thermal readings against the reference thermometer

• Calculate:

  • Mean deviation

  • Maximum deviation

  • Pass/fail rate under ±1.0°C margin

5.4 Acceptance Criteria

• ≥ 95% of thermal readings must be within ±1.0°C of the reference reading

• AI field completion must match operator input in ≥ 90% of cases

• Time stamps must align within ±1 minute of image capture

---

6. FSMS Validation Findings Template

The full validation dataset with 30 samples per CCP type is seen in the table above. Each record includes:

• Food type

• CCP category (Cooking, Cold Holding, Hot Holding)

• Reference temperature (from calibrated probe)

• Thermal image temperature

• Deviation (±°C)

• Pass/Fail result based on ±1.0°C tolerance

---

7. Risk Considerations and Controls

• Surface temperature limitation:

  This method is only valid for items where surface temperature is representative of food safety, such as:

  • Wet dishes (curries, soups)

  • Chilled foods in direct contact with air (salads, desserts)

• Not valid for internal temperatures of solid roasts or deep trays

• Operator variability:

  Validation must assess consistency across multiple users

• Environment effects:

  Account for air movement, humidity, surface moisture, and reflection

---

8. System-Level Validation

Each submission includes:

• Timestamp and device ID

• Thermal image stored with temperature data

• Location or operation context

• AI-generated food type, category, and serving conditions

These records are:

• Immutable

• Time-synced

• Searchable and exportable (PDF, CSV, Excel)

• Suitable for audit and regulatory demonstration

---

9. Conclusion & Validation Statement

Based on the above validation framework, if results meet the acceptance criteria, thermal imagery is confirmed as a valid, FSMS-compliant method for:

• Temperature monitoring at CCPs

• Auto-populated food safety records

• Supporting both real-time decision making and retrospective audit traceability

Once validated, this method is to be incorporated into the HACCP plan and reviewed annually or following any of the following:

• Hardware/software updates

• Operational scope changes

• Corrective actions resulting from deviation events

---