Telemetry that keeps working when the network does not
We fit small industrial sites — compressor rooms, cold storage, pumping stations — with sensor telemetry that survives bad links, brownouts and the occasional forklift.
Most of our work starts the same way: a site already has gauges, relays and a controller nobody has touched since commissioning, and the operator wants a number on a screen before the next inspection. We read what the existing hardware already emits rather than replacing it.
What we do
- Retrofit instrumentation
- Clamp-on current sensors, 4–20 mA loop taps and Modbus RTU reads from equipment that predates the idea of an API.
- Edge collection
- A small hardened gateway per site, buffering to local storage so a week of cellular outage costs no data.
- Alarm review
- Rewriting thresholds that fire forty times a night into ones an operator will still trust in six months.
Typical deployment
| Site | Points | Link | Buffer |
|---|---|---|---|
| Compressor room | 18–40 | LTE, wired fallback | 10 days |
| Cold storage | 60–200 | Wired | 30 days |
| Pumping station | 12–24 | LTE only | 21 days |
Commissioning a first site usually takes two visits: one to survey and pull what documentation still exists, one to mount, wire and verify against a reference meter. Everything after that we do remotely.
Working with us
We take on a handful of sites a year and keep them for a long time, so we are honest about capacity. If the timing does not work we will say so in the first call rather than the third month.
Write to hello@abs22systems.example with the site type, roughly how many points you expect, and what the network looks like on a bad day.