WiFi or Cellular, Choosing the Right Data Logger for Remote Sites

The connectivity question gets asked backwards. People ask which option is better, when the question that actually matters is what happens to your data when the connection drops.

Because it will drop. WiFi routers reboot, cellular towers get congested, and someone eventually unplugs the switch in the cupboard to charge a phone. A logger that stops recording when it loses signal is a logger that fails you on exactly the day you needed the data.

This guide covers the three connectivity options, what each one costs you in practice, and how to decide based on the site rather than the spec sheet.

Start with the failure mode, not the feature list

Every connectivity option fails sometimes. The difference between loggers is what happens next.

A logger with local storage keeps recording through an outage and uploads the backlog when the connection returns. You get a complete record with a gap in reporting, not a gap in data. A logger without it gives you a hole you can never fill.

For anything with a compliance or audit dimension, this is the whole decision. A temperature record with three hours missing from the night your chiller failed is worse than no record at all, because now you're explaining the gap.

Storage capacity in the hundreds of thousands of readings covers weeks of offline operation at typical logging intervals. Check the figure before you check the radio.

WiFi: cheapest, and fine until it isn't

WiFi is the default for good reason. No recurring cost, no SIM to manage, and setup takes minutes on a network that already exists.

It works well when the site has reliable WiFi that reaches where the sensor needs to sit, and someone technical is around when the network changes.

It causes problems when any of those stop being true. Three failure patterns come up repeatedly:

• Range. A cold room, a basement plant room or a metal-clad building will kill 2.4GHz signal at surprisingly short distances. The router is fine. The concrete is not.
• Network changes. IT rotates the WiFi password and every logger on the site goes quiet at once. Nobody notices until someone opens the dashboard a week later.
• Guest network policies. Captive portals and client isolation both break device connections in ways that look like hardware faults.

None of these is a reason to avoid WiFi. They're reasons to avoid WiFi at sites where nobody owns the network.

Cellular: independence, at a running cost

Cellular loggers carry their own connection. No dependence on site IT, no password rotations, and nothing to configure when the building's network changes.

That independence is worth real money at sites where you don't control the infrastructure: rented units, client premises, field deployments, vehicles, anywhere the alternative is asking someone else's IT department for a favour.

The trade-offs are honest ones. There's a recurring data cost, coverage has to be checked at the actual sensor position rather than assumed from a coverage map, and a SIM needs managing.

Check coverage at the sensor, not at the gate. Signal at the farm entrance tells you nothing about signal inside a metal-roofed shed at the back of the site. If you can, test with a phone at the exact mounting position before committing.

Data consumption is usually lower than people expect. Environmental readings are small, and a logger reporting every fifteen minutes moves very little data across a month. The cost driver is the connection itself, not the volume.

Ethernet and PoE: the option people forget

Where a network cable can reach, wired beats both.

Ethernet gives you the most reliable connection available, no radio interference, no password rotations, and no recurring data cost. In server rooms, plant rooms and anywhere already cabled, it's the obvious answer and it gets overlooked because people assume monitoring devices are wireless by nature.

Power over Ethernet extends this further. A PoE splitter takes an 802.3af connection and delivers both data and DC power over the single cable, which solves the two hardest problems in remote placement at once. If your constraint is that there's no outlet where the sensor needs to go, and there is a network drop, PoE is the answer.

Choosing by site type

The pattern is straightforward. Use the network when you control it, use cellular when you don't, use a cable when one exists.

Connectivity and power are one decision

These get specified separately and shouldn't be.

A cellular logger in a field needs power as well as signal, and battery life on cellular is shorter than on WiFi because the radio works harder. Solar with a battery backup is the standard remote combination for that reason.

Logging interval affects both. Reporting every minute uses more power and more data than reporting every fifteen, and for most environmental monitoring the fifteen-minute record answers the same questions. Slower intervals extend battery life and reduce data cost at the same time, so decide the interval before you size either.

One practical rule: whatever interval you choose, make sure the logger stores locally at that rate too. Reporting frequency and recording frequency are different settings, and it's the recording one that determines what you can prove later.

Choose for the site, not the spec sheet

Connectivity is the decision people spend the longest on and regret most often, usually because they chose for the building they imagined rather than the one the sensor ends up in. Test signal at the mounting position, check what the logger does offline, and match the radio to who owns the network.

Compare the options in smart sensing, or start from a configuration that's already matched in bundles where the logger and sensors are specified together.

Deploying across several sites with different network situations? Tell us what you're monitoring and where at UbiBot USA. Getting connectivity right the first time is considerably cheaper than discovering the problem after installation.

Frequently asked questions

What happens to my data if the WiFi goes down? 

On a logger with local storage, nothing. It keeps recording and uploads the backlog when the connection returns. Loggers without local storage simply lose that period, which is why storage capacity matters more than radio type for any record you might need to defend.

Is cellular monitoring expensive to run? 

There's a recurring data cost, but environmental readings are small and consumption is usually lower than people expect. The cost is in maintaining the connection rather than the data volume, and at sites where the alternative is running a cable it's normally the cheaper option.

Can I use WiFi in a cold room or freezer? 

Sometimes, but test before committing. Insulation and metal cladding attenuate 2.4GHz badly, and a signal that's strong outside the door can be unusable inside. Running the probe through the door seal and keeping the logger outside is often the better answer.

Do I need cellular if the site already has WiFi? 

Not usually, unless you don't control that network. On rented or client premises, cellular removes your dependence on someone else's IT decisions, which is worth the running cost when a password rotation would otherwise take your monitoring offline.

How much data does a monitoring device use per month? 

Very little at typical intervals. Environmental readings are a few bytes each, so a logger reporting quarter-hourly transmits a small fraction of what a phone uses in a day.

Can I change connectivity later? 

Not on the same unit. Radio type is fixed in the hardware, so the decision is made at purchase. If a site's requirements might change, cellular is the more flexible starting point because it works anywhere a network would, plus the places one doesn't exist