Why monitoring matters (and what FSANZ actually expects)

The anchor is simple: keep potentially hazardous food at 5 °C or colder (or 60 °C or hotter). When food leaves temperature control, apply the 2‑hour/4‑hour rule and record what you did. You’re also expected to show that your controls work day‑to‑day—especially under Standard 3.2.2A, which introduces practical “food safety management tools” including an evidence tool for records. In plain English: have a thermometer that’s accurate, keep sensible records (or be able to demonstrate control), and keep those records for an appropriate period.

• Temperature targets: generally ≤ 5 °C for cold foods.
• 2‑hour/4‑hour rule: a validated way to manage short excursions.
• Thermometer requirement: at least one accessible digital probe accurate to within ±1 °C.
• Evidence tool (3.2.2A): make a record for prescribed activities and keep it for ≥ 3 months, or be able to substantiate control another way.

What this means for cafés: adopt a small, reliable system (one air probe + one product‑simulator probe per cabinet, automated logging, simple daily checks). You’ll spend less time writing and more time serving customers—with better evidence when an officer asks.

Three approaches: paper, portable loggers, connected systems

There’s no one right tool. Choose based on site size, risk appetite, and how often you need to show evidence across multiple locations.

KW recommendation: start simple (one cabinet on a connected sensor) and expand once alarms are tuned. Nothing builds buy‑in like the first avoided spoilage event.

Probes & accuracy: digital, IR, air vs ‘product’

Use a digital probe thermometer accurate to within ±1 °C. Keep it accessible, clean and calibrated. Infrared “guns” are fine for a quick surface check but they don’t read the food’s internal temperature and can be tricked by shiny packaging. Equipment gauges tell you the cabinet’s air or set‑point—not the food’s temperature. For automated logging, pair an air probe (to see dynamics) with a product‑simulator probe (e.g., in a water bottle) to reflect food thermal inertia.

Where sensors go (and where they don’t)

• Air probe: in the return‑air stream, away from evaporator outlets, doors and lights. Tie‑wrap to a rack post at mid‑height; shield from direct airflow to avoid “wind chill”.
• Product‑simulator: use a small, labelled water bottle; place mid‑shelf, mid‑depth. It should ride the same thermal bumps as your real products.
• Avoid probe tips touching metal shelves, door gaskets or light fittings; they will skew readings.
• Map the cabinet: for one week, do a simple 3×3 grid of spot checks to find warm pockets. Relocate sensitive items or adjust shelving to improve flow.

Sampling rates, memory & battery life

Sampling more often doesn’t always mean better. Choose a rate that captures risk without killing batteries or filling memory.

Rule of thumb: if graphs are “hairy” but food isn’t moving, slow down sampling. If you miss short spikes during service, speed it up for those hours only.

Calibration (ice‑point), verification & drift control

Thermometers drift. That’s why you should verify at 0 °C (ice‑point) and keep a small log of corrections:

1. Fill a cup with crushed ice and a splash of cold water; stir until slushy.
2. Insert the probe to the sensing depth; avoid touching the container; wait for the reading to stabilise.
3. Record the reading and any offset. If it’s off by more than your tolerance (e.g., ±1 °C), adjust or replace.
4. Repeat monthly or after drops/repairs; note the date, probe ID and result.

For connected sensors, spot‑check against a calibrated reference probe at least quarterly and whenever data looks suspicious (flat‑lines, unrealistic jumps). Keep receipts/records of calibration as part of your evidence file.

Alarm set‑up that helps, not harasses

Alarms should tell you when to act—not force you to acknowledge pop‑ups all day. Use these patterns:

• Two thresholds: “warning” for early intervention; “critical” for action and escalation.
• Door‑open delays: allow 2–5 minutes during busy service before pinging.
• Grace periods: suppress alarms for a short time after defrost or delivery.
• Call‑tree: first to shift lead; if unacknowledged in 10 minutes, escalate to manager; then to service. Keep it short, with ownership at each step.
• Stop the noise: if a cabinet is down for maintenance, schedule a monitoring pause and add a note to the log.

KW field tip: alarms are only as good as the corrective‑action notes. Train staff to log what they did—moved stock, closed a door, called service—so you can show cause and effect in your graphs.

Records that pass audits (with templates)

Under Standard 3.2.2A’s evidence tool, keep records for at least 3 months or be able to demonstrate your control another way. Practical, audit‑friendly records include: daily fridge/freezer checks, probe calibration results, alarm summaries, corrective actions and weekly graph exports from your logger/dashboard.

Simple daily sheet (you can paste this into a spreadsheet)

Keep your sheets or PDFs for ≥ 3 months (or longer if your local regulator or corporate policy requires). Name files consistently, e.g., Store‑ID_Cabinet‑ID_YYYY‑WW.pdf.

ROI: labour saved, spoilage prevented, call‑outs reduced

Logging pays for itself three ways: fewer wasted products, fewer emergency call‑outs and less time spent writing or searching for paperwork. Use this simple math to make the case to finance:

Annual saving ≈ (Avoided spoilage) + (Avoided call‑outs) + (Labour saved)

• Avoided spoilage: one small event (e.g., A$250 of dairy/desserts) avoided every 2 months = A$1,500/year.
• Avoided call‑outs: two after‑hours visits at A$300 each the system let you prevent or schedule = A$600/year.
• Labour saved: 10 minutes/day of manual logging switched to auto + weekly manager graph export (15 min) ≈ 1.6 hours/week. At A$30/hour ≈ A$2,496/year.

Total example: A$1,500 + A$600 + A$2,496 = A$4,596/year benefit per site. Even a modest connected setup (sensors + gateway) is usually repaid in months.

Numbers are indicative. Replace with your sales mix and service rates. The point stands: one avoided spoilage + two avoided emergency call‑outs usually pays for the whole system.

Case studies (before/after curves)

Case 1 — Espresso bar: paper to Bluetooth logger

Pain: gaps in paper logs; two spoilage incidents last summer. Action: weekly Bluetooth downloads + two probes (air + product simulator) in the front merchandiser; alarm threshold with 3‑minute grace during peak. Outcome (8 weeks): false alarms down 60%; no spoilage; manager time on paperwork cut by 70% (from 50 to 15 minutes/week). Staff like seeing graphs—compliance feels concrete.

Case 2 — Dessert bar: probe placement cuts alarm noise in half

Pain: “Cabinet keeps failing” complaints every Friday night. Action: moved air probe from fan outlet to return‑air path; added product‑simulator bottle; tuned door‑open delay to 4 minutes; added a note macro for corrective actions. Outcome: alarms down 55%; better visibility of real risk; one verified door‑seal issue fixed before stock loss.

Case 3 — Multi‑site franchise: 3.2.2A evidence by design

Pain: quarterly audits required “show me” proof—paper was too patchy. Action: connected sensors on all fridges/freezers, weekly PDF exports named by store/cabinet/week, monthly thermometer ice‑point checks standardised. Outcome: consistent evidence in every store, faster audits, and fewer after‑hours call‑outs thanks to early warnings and better escalation.

Seven‑day rollout plan (start small, finish strong)

1. Day 1: pick one cabinet; install one air probe + one product‑simulator probe; set 5‑minute sampling; enable a 3‑minute door‑open delay.
2. Day 2: set up a shared folder; define file naming; create a one‑page “corrective actions” cheat sheet near the POS.
3. Day 3: draw your first week’s graph; mark defrosts and busy periods; note any false alarms.
4. Day 4: run an ice‑point check on the handheld probe; record the result; replace batteries if needed.
5. Day 5: train the team: where probes live, what an alarm means, how to log actions in one sentence.
6. Day 6: decide whether to roll out to the next cabinet; set sampling slower (10–15 minutes) on back‑of‑house storage to save battery.
7. Day 7: export your first weekly PDF; sign and file; book the next review in your calendar.

Commissioning note for NEW units: when a new fridge/freezer arrives, let it stand upright with power off for 6–8 hours (12–24 hours if it travelled on its side), then power on and allow a full pull‑down before loading. Log the first stable readings. This single habit prevents weeks of nuisance faults.

FAQ

How often should I log temperatures?

For storage units, 10–15 minutes is usually fine; for high‑traffic displays, 1–5 minutes. Always back your choice with stable graphs and fewer alarms, not just theory.

Do I need two probes per cabinet?

Not required, but highly useful. The air probe shows dynamics (doors/defrost), while a product‑simulator tells you what your food experienced. Together, they make audits easier and reduce false positives.

What thermometer should I buy?

A simple digital probe accurate to within ±1 °C meets requirements. Keep it accessible, sanitise before/after use, and check it monthly at the ice‑point.

How long must I keep records?

At least 3 months under the Standard 3.2.2A evidence tool, unless you can substantiate controls to an authorised officer in another acceptable way. Many chains keep 12 months for comfort.

Can I rely on the cabinet’s display?

Cabinet displays are useful but often show air or set‑point. For proof, measure food with a probe or use a product‑simulator in your logger.

What about energy?

Better logging reduces door‑open drift and catches ventilation or gasket issues early, which in turn lowers compressor run‑time. It’s the cheapest “control upgrade” you can make.

References

• FSANZ — Keeping food at the right temperature (≤ 5 °C) and 2‑hour/4‑hour rule.
• FSANZ — Thermometers (accessible probe; ±1 °C accuracy; digital probe vs IR; cleaning & calibration).
• FSANZ — Standard 3.2.2A overview and the Evidence tool (PDF) (records ≥ 3 months; alternatives to records).
• NSW Food Authority — Showing food is safe and Record‑keeping templates.

We summarise public guidance for operators. Always follow your model’s manual and the current instructions from your local regulator.

Related guides

• Commercial Fridge Installation & Ventilation — clearances & site‑survey
• EEI under GEMS 2024 — check registration & estimate running cost
• Stop the Ice — defrost, airflow & evaporator hygiene
• Ask us to design your logging & alarm setup