What are the ways to monitor a refrigeration system?
Each option relies on a different signal. Some only measure temperature, some read the system’s own controller, and some measure the refrigeration system itself. That decides what they can detect, and how early.
Data logger or temperature logger
A probe that measures and records the temperature in a cold room, a freezer room or a fridge. Many send their readings wirelessly to an app and can raise an SMS alarm.
Controller alarm
The electronic controller on the system has built-in alarms for high temperature, high and low pressure and probe faults. The alarm shows on the display and can be wired on to a sounder, a relay or a GSM dialler.
Manned alarm monitoring centre
A centre staffed around the clock receives alarms from the system and phones the people on an agreed call list.
BMS monitoring
The building management system pulls values and alarms from the refrigeration controllers, usually over Modbus or BACnet, and shows them alongside ventilation, heating and power.
Predictive monitoring (Maci)
Dedicated sensors measure current, pressure and temperature on the system. AI learns how each system normally runs and warns you when a signal starts to drift.
What does each option detect, and how early?
The biggest difference between the options is what they measure. If an option only measures temperature, or reads the controller’s alarms, it can only tell you once the fault has shown itself. If it measures the system itself, it can see the fault developing.
| Option | Measures | Detects | How early |
|---|---|---|---|
| Data logger | Room temperature | Temperature too high or too low | Once the temperature has crossed the limit, often hours after the cooling has failed |
| Controller alarm | What the controller itself measures, usually temperature and pressure | High temperature, high and low pressure, probe faults, stop | Once a limit is crossed or the system has stopped |
| Manned monitoring centre | Nothing itself. Receives alarms from a controller or logger | The alarms that are sent in | At the same moment as the alarm, plus the time it takes to reach someone |
| BMS monitoring | The values the controller exposes | Alarms and limits crossed | Once a limit is crossed. Slow changes only show if someone watches the trend graphs |
| Predictive monitoring (Maci) | Current, pressure and temperature from its own sensors, every minute | Drift from the system’s normal behaviour, with the likely cause | Typically 3 to 6 weeks before a breakdown, for faults that develop gradually |
Timing matters. An alarm that comes when the room is already warm leaves hours to save the stock, and often means a technician called out at night. In Denmark, a typical emergency call-out is about 15 minutes of work and costs about €360 excluding VAT. A warning weeks ahead leaves time to order the part and fix the fault on a planned daytime visit. If the system has already stopped, read what to do in a refrigeration breakdown.
What does each option need, what are the drawbacks, and who is it for?
All five options have their place. The table shows what each one asks of you, where it falls short, and where it fits best.
| Option | Needs | Drawbacks | Best for |
|---|---|---|---|
| Data logger | A probe in each room, battery changes, and often a gateway and an app subscription | Temperature only, no cause, and a late alarm | Small sites with a few cold rooms or fridges, and for HACCP records |
| Controller alarm | Nothing extra, as long as the alarm is wired to something that reaches a person | Fixed limits and late alarms. If it only shows on the display, nobody sees it at night | Every system. It should always stay switched on as a safety net |
| Manned monitoring centre | A contract with a centre and an alarm signal from the system | A running cost per connection, and the same late signal | Sites without their own on-call rota, where a person should judge the alarm and ring round |
| BMS monitoring | A BMS, plus integration and programming for each controller | Depends on the make and on what the controller exposes. Fixed limits | Large buildings that already have a BMS and an in-house technical team |
| Predictive monitoring (Maci) | Sensors fitted by a certified Maci partner, and a wired network or an LTE modem | Needs installing, and Maci first learns the system. Systems with a compressor only. Not manned | Shops, cold stores, chains and FM companies, where breakdowns and call-outs are costly |
What is the difference between a data logger and refrigeration monitoring?
A temperature data logger measures the temperature at set intervals and stores it. Modern loggers send their readings wirelessly to an app and raise an SMS alarm when the temperature goes out of range. They are cheap, simple to set up and good for keeping HACCP temperature records.
The limitation is that a logger only sees the room. By the time the temperature rises, the fault in the system has already happened, and the logger cannot tell you whether it is the compressor, a leak or a door left open. Monitoring the refrigeration system itself also measures current and pressure, and can see the fault before the temperature rises. Read more about monitoring the whole system.
If you have a few cold rooms and mainly need records, a logger may be enough. Read more about HACCP temperature monitoring.
Is the controller alarm enough?
The controller alarm is already there, and it should always stay switched on. It reacts to high temperature, high and low pressure and probe faults, and the controller can stop the compressor to protect it.
The controller alarm is only enough if it reaches a person. If it only shows on the display, someone may not notice it until the next morning. The limits are fixed and set wide enough that normal days do not trip them, so the alarm comes once the fault is already there. It is usually this alarm that turns into a night-time call-out.
Many sites therefore combine the two: the controller alarm as a safety net, and predictive maintenance to catch faults before they trip it.
What do you get from a manned alarm monitoring centre?
A manned alarm monitoring centre is staffed around the clock. When an alarm comes in, an operator phones the first person on the list and, if there is no answer, the next. The centre can judge the alarm, log who has taken it on, and call out the technician.
That is a real advantage if you have no on-call rota of your own. A monitoring centre can only react to the signals it receives, though. If the signal comes from the controller or a logger, the fault has happened by the time the centre hears about it. The price is usually a fixed monthly fee per connected system or site.
Maci Intelligence is not a manned monitoring centre. Maci watches around the clock, sends alerts in the app, by SMS and by email, and escalates automatically to the next person on the list if nobody responds. Your refrigeration contractor can be on the list as the on-call contact. Read more about refrigeration alarm monitoring.
What is BMS or Modbus monitoring, and what can it not do?
A building management system (BMS) is the building’s shared system for ventilation, heating, power and often refrigeration. The refrigeration controllers are connected over a protocol such as Modbus or BACnet, and the BMS shows their values and alarms on the same screen as the rest of the building.
The advantage of BMS monitoring is one overview for the site team, and the ability to control the plant and change set points. The limitation is that a BMS only sees what the controller exposes, and that depends on the make, the model and how the integration was programmed. Many controllers expose temperatures and alarms, but rarely the compressor’s current draw. Alarms are based on fixed limits, and slow changes are only noticed if someone follows the trend graphs.
Maci Intelligence reads nothing from the controller and measures with its own sensors, so the readings are the same on every system. The data can be pulled into other systems through Maci’s API.
What does predictive monitoring detect that an alarm does not?
Predictive monitoring looks at how hard the system is working. Maci Intelligence measures current, pressure and temperature at several points on the system every minute, learns how each system normally runs, and warns you when a signal starts to move away from normal. The alert says what is drifting, the likely cause, and how long it is likely to be before it becomes critical.
- A wearing compressor: current draw and discharge temperature climb over several weeks.
- A refrigerant leak: suction pressure falls slowly, and the compressor runs for longer.
- A dirty condenser or a failing fan: condensing pressure and current rise.
- An evaporator icing up or a door seal leaking: longer defrosts and bigger swings in the room.
Predictive monitoring has its limits too. The sensors have to be fitted by a certified Maci partner, and Maci first has to learn how the system normally runs before the warnings start. Maci monitors systems with a compressor. Sudden faults such as a power cut, a tripped breaker or a door left open give no warning weeks ahead. Maci is not manned, so someone on the contact list has to act on the alert.
Read more about refrigeration monitoring with Maci and about how the AI sees a breakdown coming.
Can the options be combined?
Yes, and that is the most common set-up. The controller alarm should always stay switched on. Predictive monitoring can be added on top without changing anything on the system, because Maci neither reads from the controller nor controls the system.
- Controller alarm and Maci: the safety net stays, and faults that develop gradually are seen weeks ahead.
- Data logger and Maci: Maci also measures air and product temperature. Whether Maci’s data can replace the logger in your HACCP records is for your business to judge.
- Monitoring centre and Maci: when more faults are fixed on planned visits, there are fewer urgent alarms to handle. Your contractor can also go straight onto Maci’s alert list.
- BMS and Maci: the BMS runs the building, and Maci watches the compressor and the refrigeration circuit, and its alerts can be pulled through the API or sent to the BMS with a webhook.
What should you choose for one cold room, and what for many sites?
The choice depends on how much a breakdown costs you, and on how many systems you have. Here is a rule of thumb.
| Your situation | Usual choice |
|---|---|
| One cold room or a few fridges, mainly needing records | A data logger with SMS alarm, plus the controller’s own alarm |
| A shop, a commercial kitchen or a cold store where a breakdown costs stock and sales | The controller alarm plus predictive monitoring of the most important systems |
| A building with a BMS and technical staff | The BMS for the building, plus predictive monitoring of the refrigeration systems |
| No on-call rota of your own at night | A manned monitoring centre, or your contractor’s on-call technician on the alert list |
| Many sites, a chain, or an FM company running clients’ systems | Predictive monitoring of every system, with one overview and escalation to your contractor |
If you run refrigeration systems for clients, see refrigeration monitoring for facility management.
How much do the different options cost?
The options are priced in different ways, and the monitoring itself is rarely the biggest cost. The faults it fails to prevent are.
| Option | Usual pricing |
|---|---|
| Data logger | Buying the probes and perhaps a gateway, often with a subscription for the app and data storage |
| Controller alarm | No extra cost beyond the controller already fitted. The cost comes when the alarm turns into a call-out |
| Manned monitoring centre | A fixed monthly fee per connected system or site, plus call-outs |
| BMS monitoring | A project price for integration and programming, plus running and maintaining the BMS |
| Predictive monitoring (Maci) | The price depends on the number of systems and sensors, and is given as a quote after a short review |
Set the price against what faults cost today: in Denmark a typical call-out costs about €360 excluding VAT, before parts ordered in a hurry and stock that has to be thrown away. Read what a refrigeration breakdown costs and the guide to choosing a monitoring system.
Questions and answers
Which option finds a fault earliest?
Predictive monitoring with sensors on the system itself. Maci Intelligence warns typically 3 to 6 weeks before a breakdown for faults that develop gradually, such as compressor wear, refrigerant leaks and a dirty condenser. A data logger, a controller alarm, a monitoring centre and a BMS usually tell you once the fault has happened.
Can Maci replace an alarm monitoring centre?
Maci Intelligence watches around the clock and escalates alerts automatically to the next person on the list. Maci is not manned. If you need a person to judge the alarm and ring round, you can keep the monitoring centre, or put your contractor’s on-call technician on Maci’s alert list.
Should the controller alarm be switched off once Maci is fitted?
No. The controller alarm is the system’s own safety net and should always stay switched on. Maci does not read from the controller and does not affect its alarms or settings.
Can a BMS do the same as predictive monitoring?
A BMS can show trends and raise alarms at fixed limits, but it only sees what the controller exposes and rarely gets the compressor’s current draw. Slow changes are only caught if someone follows the graphs. Predictive monitoring watches for them automatically on every system.
Is a data logger enough for HACCP?
For many small sites, a data logger can keep the temperature records that HACCP calls for. What your own food safety plan requires is for your business to judge. Read more about HACCP temperature monitoring.
How much does predictive monitoring from Maci cost?
The price of Maci Intelligence depends on the number of systems and sensors, and you get a quote after a short review of your systems. Maci Service, with QR tags and service history, is free.