Telemetry and online sensors in monitoring - automatic data recording
Classic monitoring means a trip into the field for every reading. Telemetry changes this logic: the sensor stays in place, measures continuously, and the data flows in by itself - to the cloud or to a platform. For many applications this is a huge leap in quality, because instead of a handful of points you have a dense, continuous series.
How it works
- A sensor / probe (often multi-parameter) measures at a fixed interval and records the data in a logger.
- Telemetry (e.g. by a GSM/LoRa module) transmits the readings live - without waiting for a visit.
- The data reaches a platform, where it forms charts, and once a threshold is breached - alerts.
The chain has four links and each fails differently - and each failure leaves a different trace in the data. This is practically the whole diagnostics of telemetry:
| Link | What it does | Typical failure | How it looks in the data |
|---|---|---|---|
| Sensor / probe | measures at a fixed interval | drift, biofouling, dirty optics | slow departure from a reference measurement, a “creeping” baseline |
| Logger | stores locally | full memory, wrongly set clock | gaps in the series, or measurements shifted in time |
| Link (GSM / LoRa) | transmits the readings | no coverage, flat battery | the series stops entirely, and after repair the data arrive in a lump |
| Platform | receives, plots, alerts | wrong time zone, duplicates, interrupted polling schedule | a day shifted by hours, doubled points, silent silence instead of an error |
⚠️ The most dangerous telemetry failure does not look like a failure. A drifting or fouled probe goes on sending nice, dense data - only untrue. That is why a periodic reference measurement with a handheld instrument is part of the method, not a luxury, and why a break in the inflow of data should raise an alert just as an exceeded threshold does.
What you gain
- High resolution - you can see daily fluctuations and short events that spot measurement misses.
- Early warning - a threshold + a notification lets you react in time (e.g. a fall in oxygen, the temperature approaching 0°C, a spike in conductivity).
- Fewer trips - the sensor works on its own between servicing.
What to keep in mind
- Calibration and servicing - sensors drift and get fouled; without maintenance the data quietly loses its credibility.
- Power and connectivity - in the field these are real constraints (battery, signal).
- Quality control - it is worth periodically comparing the sensor with a reference measurement.
📊 See it live: Reef & Coast Watch - a public dashboard from six NOAA tide gauges (Hawaii, Guam, Florida, US Virgin Islands, Puerto Rico): water temperature and water level fetched automatically, with a station map. No login.
Sources
- Running a continuous-monitoring station - calibration, drift correction, servicing and record reporting: Wagner R.J., Boulger R.W., Oblinger C.J., Smith B.A. (2006), Guidelines and Standard Procedures for Continuous Water-Quality Monitors: Station Operation, Record Computation, and Data Reporting, U.S. Geological Survey, Techniques and Methods 1-D3 - pubs.usgs.gov. It supersedes the earlier report WRIR 00-4252.
(Addresses checked on 18 August 2026.)
In practice
Telemetry only makes sense when the data arrives in one place by itself and can be read together with context. LimnoLog takes in measurements from sensors through connectors / telemetry and records them in your project - alongside manually entered data and public data (weather, hydrology) on a shared time axis. A warning threshold on an indicator triggers a notification. This turns a continuous stream of data into real supervision, not just an archive of figures.
See it in the LimnoLog app
Explore LimnoLog features