Where should you place temperature sensors in a cold room? A complete guide to temperature mapping

Plenty of cold rooms have modern temperature monitoring systems and still lose stock. The cause is rarely the equipment — it is usually where the sensor is placed. A sensor mounted right under the evaporator will always show a nice number, while cartons in the corner by the door or up near the ceiling may have been out of range for hours.

This article explains how to do cold room temperature mapping the way GSP-certified pharmaceutical warehouses do it, simplified so that owners of food, produce and seafood cold rooms can do it themselves with a handful of wireless sensors.

Why does temperature vary so much inside one cold room?

Cold air does not spread evenly on its own. It follows the path blown by the evaporator fans, hits the stock and returns to the evaporator. Anywhere that airflow barely reaches will be warmer. In practice, a 3–6 °C difference between two points in the same room is normal. The most common causes:

• The door: every time it opens, warm humid air pours in along the floor and spreads down the aisle.
• Near the ceiling, far from the evaporator: warm air rises, and if the fans do not reach that area it warms up first.
• Sun-facing walls, metal roofs, spots near lights or where pipes pass through.
• Tightly stacked stock: pallets pushed against walls or the ceiling, or blocking the air outlet, choke the airflow.
• Directly in front of the evaporator outlet: this is the opposite — a cold spot. Vegetables and fruit placed here can suffer chilling injury or freeze.

The refrigeration controller's sensor usually sits in the return air to the evaporator. It tells you how the refrigeration unit is working, not what temperature your products are at. That is why you need a separate monitoring system, placed in the right spots.

What is cold room temperature mapping?

Mapping means placing many temporary sensors in a three-dimensional grid inside the room, logging continuously for several days, then analysing which points are hottest and coldest and whether the room holds the required temperature at every position. The World Health Organization's technical guidance on medicine storage (the supplement on temperature mapping of storage areas) is the most widely used reference.

Mapping answers three questions:

1. Does the room meet the temperature requirement at every position where stock is stored?
2. Where are the hot spot and cold spot — the places to put permanent monitoring sensors?
3. Are there areas where stock should not be stored (for example right in front of the air outlet or next to the door)?

Preparation: how many sensors?

There is no fixed number for every room, but this method works for most small and medium cold rooms:

• Floor plan: divide the floor into a grid with cells about 5–10 m across (rooms under 50 m² can use 3–4 m). Each intersection is a "measurement column".
• Height: each column gets sensors at 3 heights — low (about 30 cm above the floor), middle, and high (level with the highest stock). Rooms lower than 2.5 m may need only 2 heights.
• Mandatory extra points: next to the door, in front of the evaporator outlet, the corner furthest from the evaporator, next to a sun-facing wall, and 1 sensor outside the room to compare with ambient temperature.

Room area	Height	Suggested sensors
Under 20 m² (mini room, container)	≤ 2.5 m	6–9
20–100 m²	2.5–4 m	9–15
100–500 m²	4–8 m	15–30
Over 500 m²	High-rack storage	Plan a dedicated grid

Mapping sensors must be checked against a reference thermometer before use. A simple method: put all sensors side by side in one stable place for 1–2 hours, record each sensor's offset, and correct for it during analysis.

How long and when should you measure?

Mapping must reflect how the room is actually used every day. So:

• Measure for at least 3 consecutive days, ideally 7 days to include both working days and days off.
• Include the peak loading and unloading hours (frequent door openings) and the evaporator defrost cycles.
• Log every 1–5 minutes. Longer intervals miss short temperature spikes when the door opens.
• If the room runs year-round in a hot climate, map again in the hottest season.
• A full room and an empty room behave differently. If possible, measure with the room loaded as usual.

This is where wireless sensors shine. Instead of data loggers that you have to collect and plug into a computer one by one, LoRa nodes send data to the gateway continuously. You watch charts for every measurement point on your phone during the study and can spot anomalies on day one instead of waiting until the end.

Reading the results: finding hot and cold spots

Once you have the data, calculate 4 numbers for each sensor: minimum, maximum, average and time out of range (in minutes). Then:

1. The hot spot is the sensor with the highest maximum and the longest time above the upper limit. Usually near the door or near the ceiling far from the evaporator.
2. The cold spot is the sensor with the lowest minimum. Usually right in front of the air outlet.
3. Look at the recovery time after each door opening: how long each point takes to get back below the limit. If it takes more than 30 minutes, review the stacking pattern or install an air curtain or strip curtain.
4. Look at the effect of defrosting: how much the temperature rises and for how long. Use this to set alarm delays and avoid false alarms.

Draw the results on the room's floor plan (one colour per point based on its maximum temperature). A simple picture like that convinces management and inspectors far better than a table of numbers.

Where should permanent monitoring sensors go?

Only after mapping do you know where to put long-term sensors. General rules:

• At least 1 sensor at the hottest spot — the place that alarms first when something goes wrong.
• 1 sensor at the coldest spot if your products are sensitive to cold (vegetables, tropical fruit, vaccines, milk).
• 1 representative sensor in the middle of the stock to know the "real" product temperature.
• 1 door switch to know how long the door stays open — it explains temperature spikes.
• Do not let probes touch walls, the floor or the evaporator, or sit right in front of the air outlet (unless that is the point of the measurement).

Tip: use a "thermally buffered" probe to cut false alarms

A bare probe measures air temperature, so every time the door opens for a few seconds the reading jumps even though the products barely change. Pharmaceutical warehouses commonly put the probe inside a small vial of glycol solution or dry sand (a buffered probe). The buffer reacts slowly like a real product, so the reading is much closer to product temperature and false alarms from quick door openings almost disappear. For food, you can use one bare probe (early warning) and one buffered probe (product quality assessment).

When should you map again?

• After replacing or substantially repairing the evaporator, compressor or fans.
• After changing the rack layout or increasing stacking height.
• After changing the room's purpose (for example from 2–8 °C chilled storage to frozen storage).
• After a major temperature incident with an unclear cause.
• Periodically — pharmaceutical warehouses usually repeat it every year or every season.

How to map with the AIoT Viet kit

The cold-storage monitoring solution is designed for both jobs: the initial study and long-term monitoring.

1. Install 1 gateway outside the room (WiFi with 4G backup).
2. Mount STM32 or ESP32 nodes with external temperature probes following the grid above. Each node also has an outside temperature–humidity sensor and a door switch.
3. Set the reporting interval to 1–2 minutes during mapping and watch everything live on the dashboard.
4. When the study ends, export the data to Excel for analysis, keep the nodes at the hot spot, cold spot and representative point, and change the interval back to 5 minutes to save battery.

This is much cheaper than hiring a mapping contractor every year, and you keep all the data for future inspections.

Frequently asked questions

Does a small 10 m² cold room need mapping?

It should be done, but on a small scale: 6 points (3 corners × 2 heights) for 3 days is enough to see whether any area is unusually warm. Small rooms with frequent door openings often have even larger differences.

Can cheap digital thermometers be used for mapping?

Yes, if they log data and you have checked their offset against a reference thermometer. Display-only thermometers are no use, because you will have no night-time data.

Do sensors inside the cold room need to be waterproof?

Yes. Cold rooms are humid and water drips during defrosting. Probes should be stainless-steel sheathed and waterproof, with the electronics and battery in a sealed enclosure — for freezer rooms, keep the enclosure outside and run only the probe inside.

Can the mapping results be submitted to inspectors?

A mapping report (sensor location plan, study period, data, conclusions) is a useful document for HACCP and GSP inspections. For pharmaceutical storage, also check the specific requirements of your regulator and the calibration certificates of your instruments.

Want to map your own cold room? See the kits in the cold-storage monitoring solution or use the kit selector to find out how many nodes your room size needs.