This is the multi-page printable view of this section. Click here to print.

Return to the regular view of this page.

Blog — Off-Grid Tech & OGLAS Updates

Articles on off-grid sensor networks, long-range wireless monitoring, and OGLAS development. Practical guides for anyone building resilient, no-subscription monitoring — on farms, fleets, factories, or remote sites.

Stories, guides, and updates from building off-grid sensor networks. Many of the examples come from farms — our test bed — but the ideas apply anywhere off-grid.

Why Your Farm Data Should Stay on Your Farm

The case for local-first farm sensor data — no cloud accounts, no monthly subscriptions, no vendor lock-in. Your sensor readings, your storage, your formats, forever.

Every smart-farm product on the market wants your data on their servers. Not because it’s better for you — because it’s better for their recurring revenue model.

Here’s why we built OGLAS to work the other way.

The subscription trap

A typical cloud-connected farm sensor setup:

  • $5–15/month per device for connectivity (SIM plan, API access, “premium features”)
  • Data held hostage — cancel the subscription, lose access to your history
  • Feature tiers — want alerts? That’s the Pro plan. Want more than 90 days of history? Enterprise.
  • End-of-life risk — company gets acquired or shuts down, your hardware is e-waste

Over a ten-year farm deployment — and farm infrastructure should last a decade — the subscription costs dwarf the hardware by an order of magnitude.

What “local data” actually means

With OGLAS, the hub logs every sensor reading to local storage — flash, SD card, a NAS, your choice. The format is plain text and easily-parsed binary. There’s no mandatory OGLAS cloud account — local is the default, and nothing is pushed off your site unless you choose it.

This means:

  • No subscription. Buy the hardware, run it, done.
  • No lock-in. In five years, if you want to move to a different system, your data is in plain formats. You’re not negotiating an export from someone’s API.
  • No account. No login, no password, no app-store dependency.
  • No telemetry. The hub doesn’t phone home. Updates are pulled when you choose.

The trade-off, honestly

You are responsible for your own backups. We can help you set them up — SD card redundancy, rsync to a NAS, whatever suits your setup — but nobody else is going to do it for you. That’s the deal: total control comes with total responsibility for the data.

We think that’s a good deal. If the power goes out, your sensors still talk to your hub. If the internet goes out, your hub still logs. If we disappear as a company, your hardware still works, your data is still yours, and the formats are open and documented.

What about remote access?

“Local” doesn’t mean “can’t access it remotely.” It means the default is local, and remote access is a choice you make:

  • Tailscale or WireGuard into your home network — secure, no open ports
  • A VPS you control running a dashboard that pulls from your hub (or the optional OGLAS Cloud, on your terms)
  • A phone app on the local network when you’re in range

All of these are your infrastructure, on your terms. No third party sits between you and your data.

The industry is starting to notice

The “local-first” movement isn’t just us. From Home Assistant in home automation to ChirpStack in LoRaWAN networks, the pattern is the same: people who run infrastructure on their own property want control over the data that infrastructure generates.

OGLAS is that idea applied to farm sensors.


This is pillar one of Why OGLAS. The other three: no telco dependency, monitoring and alerts, and off-grid power.

How to Monitor Remote Water Troughs Without Driving the Paddocks Every Day

GPS-pinned LoRa water trough sensors that report level, alert on low water, and run for months on battery. Stop doing the daily trough run by ute.

If your property has twenty troughs spread across a few thousand hectares, checking them all is a half-day job — and it’s a half-day you repeat every day because the cost of missing a dry trough is measured in stock weight, stress, and mortality.

Here’s how we designed the OGLAS Water Trough Level sensor to make that drive optional.

The problem with “just check them daily”

  • Time: Twenty troughs at five minutes each (including the drive between paddocks) is nearly two hours a day. Over a year, that’s 700+ hours — basically a month of full-time work.
  • Reaction time: If a trough runs dry at 10am and you check it at 4pm, the stock have been without water for six hours. On a hot day, that’s a welfare issue.
  • False security: “I checked them all this morning” becomes a reason not to worry — right up until the float valve sticks at 10:30am.

What the sensor does

  • Reports level periodically — default once an hour, configurable down to minutes during commissioning
  • Raises an immediate out-of-water alert — independent of the reporting interval, this message does not wait
  • Includes GPS coordinates with every reading, so the hub knows which trough is low without you maintaining a node-name-to-paddock spreadsheet
  • Reports its own battery voltage so you replace before failure

Why GPS matters

Twenty troughs means twenty node names, twenty positions, and twenty entries in a spreadsheet somewhere that nobody updates when a trough gets moved. Miss one update and “trough-7 is low” becomes a guessing game.

With GPS, the trough self-reports its location every cycle. The hub plots the troughs on a map. “The trough at -38.51, 145.20 has dropped to 80 mm” — actionable without a paper map.

Battery life: the design constraint

Remote troughs are usually a long way from power. The sensor is built around battery life:

  • Ultra-low-power hardware for minimal deep-sleep current
  • Deep sleep between cycles — GPS, radio, and level sensor fully powered down
  • Confirm-then-sleep — wake, read, get a GPS fix, send, wait briefly for confirmation, sleep
  • Under 30 seconds of active time per hour

A single 2000 mAh cell with a small solar trickle runs year-round.

Pair it with a Smart Switch

If a trough drops below threshold, the Smart Switch at the pump turns on and fills it — automatically. The sensor reports the low; the switch acts on it; the hub logs both. No human in the loop for a routine refill.


Water trough sensors are in the OGLAS sensor catalogue. Get in touch if you’ve got a property-full of troughs you’d rather not drive to every day.

Building an Electric Fence Monitor That Alerts You the Moment the Fence Dies

How to build a battery-powered LoRa sensor that checks your electric fence is pulsing, alerts you when it stops, and runs for a full season on a single charge.

An electric fence that’s stopped pulsing is just a fence — and livestock work that out fast. The problem isn’t detecting a dead fence; it’s detecting it before you do the morning rounds and find animals where they shouldn’t be.

Here’s the design behind the OGLAS Electric Fence Active sensor — and how you can build one yourself.

The detection principle

You don’t need to wire into the high-voltage side. A small antenna held a few centimetres from the fence wire picks up the induced voltage from each pulse via capacitive coupling. Every pulse triggers a counter. No pulse for N seconds? The fence is down.

This is safer (no galvanic connection to kilovolts), easier to install (zip-tie to a fence post), and immune to the energiser’s voltage level — it either pulses or it doesn’t.

What you need

  • Board: Ultra-low-power wireless board — chosen for its minimal deep-sleep current
  • Pickup: A few centimetres of wire as an antenna, heatshrinked and pointed at the fence line
  • Power: Single 18650 or LiPo cell, with a small solar trickle panel
  • Enclosure: IP65 sealed box, magnetically or zip-tied to the post

Why multiple sensors matter

A single sensor at the energiser tells you the charger is working. It doesn’t tell you the far end of the line still has voltage.

For long boundaries, deploy two or three:

Position Catches
At the charger Charger / battery faults
Quarter-way out Breaks in the first half
Far end Breaks anywhere, line voltage drop

Each sensor reports independently with its own node name (fence-east, fence-far), so you know which segment failed — and which paddock to walk to.

Alert immediately, not on schedule

Most sensor systems report on a schedule. “Check in every 5 minutes” means your fence could be down for 4 minutes and 59 seconds before you find out. That’s long enough for cattle to notice.

The OGLAS electric fence sensor reports on schedule and immediately on state change. Fence drops from active to inactive? Message fires right now, not at the next interval. The bell in the homestead rings the moment the alert lands.

Battery life that matters

The sensor sleeps between pulses. Wake, check for a pulse, update the counter, sleep. Full power-down of the radio and the GPS between cycles. The whole active cycle is under a second per check.

On ultra-low-power hardware with a 2000 mAh cell and a small solar panel, it’ll run for a full season without attention — and reports its own battery voltage so you can swap before it dies.


OGLAS electric fence sensors are part of the sensor catalogue. If you’ve got a long boundary and a story about stock getting through, we’d like to hear it.

LoRa vs 4G for Farm Sensor Networks: Which Actually Costs Less Over 5 Years?

A real cost comparison between LoRa-based off-grid farm sensors and 4G-connected alternatives — hardware, SIM plans, and hidden costs over a 5-year deployment.

When you’re putting sensors across a farm — gates, water troughs, tanks, fence lines — every device that needs a SIM card is a device with a monthly bill. Ten sensors at $5/month is $600/year before you’ve read a single tank level. Over five years, that’s $3,000 in connectivity alone.

LoRa changes the arithmetic. Here’s the real cost comparison.

The hardware

LoRa (OGLAS) 4G/NB-IoT sensor
Sensor board $15–30 (readily available LoRa board) $40–80 (SIM7000-based)
Hub $30 (low-cost wireless board) None (each device phones home)
SIM card None $2–5/month per device
Enclosure + power $10–20 $10–20
Per-sensor cost, year 1 $55–80 $100–160

The connectivity

Here’s where it gets ugly for the 4G approach.

Ten sensors on a $5/month IoT SIM plan:

  • Year 1: $600
  • Year 5: $3,000 cumulative
  • Year 10: $6,000 cumulative

The LoRa hub? One low-cost wireless board. No SIM. No plan. The sensors talk to it over LoRa, and the hub is the only device that needs internet — and only if you want remote access.

The hidden costs

4G coverage. Most farms have dead spots. That back paddock with the trough? No signal. You’re either moving the sensor closer to the tower (and further from the thing you’re monitoring) or paying for a repeater.

SIM management. Ten SIMs means ten accounts to manage, ten renewals to track, ten “why is this one over its data cap?” investigations. It’s administrative overhead dressed as a product.

End-of-life risk. 4G modems get discontinued. 2G and 3G shutdowns have already bricked thousands of agricultural sensors. LoRa is a physical layer — the radio doesn’t care what network generation we’re on.

The LoRa trade-off

LoRa isn’t magic. You need a hub within range (typically 2–10 km depending on terrain). You’re responsible for the infrastructure. But on a farm, you already own the infrastructure — the shed, the pole, the solar panel. Adding a $30 hub is a one-time cost.

Five-year total: 10 sensors

LoRa 4G
Hardware $750 $1,100
Connectivity $0 $3,000
Total $750 $4,100

The gap widens with every sensor you add and every year you run them. That’s not a pricing strategy — it’s physics. LoRa is your radio, on your property, carrying your data.


We build OGLAS to make this choice easy. Sensors talk LoRa. The hub lives in your shed. No subscriptions, no telco, your data stays local. Browse sensors →