Monitor and control your equipment remotelyA controller in a vineyard plot, a machine in a workshop, a sensor at the end of a field: see, understand, act.

Your equipment already produces data: a Niagara or Siemens controller, a meter, a weather probe, a pump. It often stays locked in the vendor's tool, slow and reachable from a single workstation. We bring it into a platform your teams open on their phone, with history, alerts and, where relevant, remote control.

Your situations

What we meet in the field

You have to travel on site to know the state of an installation.

The vendor's tool takes forty-five seconds to answer and only works from a dedicated PC.

Alarms exist on the controller, but nobody receives them remotely.

Readings are written by hand, then copied into a spreadsheet.

Three equipment brands, three software tools, no overall view.

The site has no coverage, or a 4G link with a limited plan.

What we deliver

What we build

The gateway to your equipment

A service that speaks your controllers' and sensors' protocol, BACnet/IP, Modbus, OPC UA, MQTT, LoRaWAN, and brings data up without altering it or bypassing their safeties.

The monitoring platform

History in a time-series database, dashboards per site and per device, a web app installable on mobile, accounts per role and per organisation.

Alerts and control

Thresholds, missing readings, faults: notifications on phone and by e-mail. Remote commands when the equipment allows it, with local safeties always taking priority.

Operations

Hosting in France, tested backups, updates, monitoring of the platform itself: we keep running what we deliver.

How it works

From controller to application, in four steps

  1. 01

    Communication driver

    We connect to a real device, read its points, measure. That is where we discover the quirks of the protocol and the vendor.

  2. 02

    Data model

    What we measure, how often, how long we keep it, what triggers an alert. Written with the people who run the installation.

  3. 03

    Platform and application

    Collection, time-series database, web and mobile interface, alerts, accounts. A first usable version on one site before extending.

  4. 04

    Rollout and operations

    Go-live, training, then extension to other sites. The platform is monitored like any critical application.

A delivered example: vineyard covers controlled remotely

A manufacturer of retractable vineyard covers wanted to offer its customers monitoring of their plots. Each plot is driven by a Niagara Edge 10 controller, reachable on site or through the vendor tool, at forty-five seconds per request.

We wrote a BACnet/IP driver that reads the weather sensors, the thresholds, the position and fault code of each row, and sends deploy and retract commands, in under two seconds. On top, a web and mobile platform: history in a time-series database, alerts, four permission levels, multi-estate. The controller's automatic safeties, frost, hail, rain, wind, always take priority over a manual command.

The protocols we speak

  • BACnet/IP: building and equipment controllers, Tridium Niagara, Distech, Siemens.
  • Modbus TCP and RTU: meters, drives, industrial PLCs.
  • OPC UA and MQTT: gateways and recent devices.
  • LoRaWAN: low-power sensors over large areas, with a private ChirpStack network where public coverage is missing.
  • 4G and constrained links: outage-tolerant collection, deferred upload, controlled data usage.

What separates useful monitoring from one more dashboard

  • Data is reliable: every point carries its timestamp and quality; missing data is information, not a hole.
  • Alerts are useful: few, calibrated with operators, each with an expected action.
  • Control is safe: never by bypassing the equipment's safeties, always logged, with per-person rights.
  • It works in the field: on a phone, with a poor connection, by someone wearing gloves.

And where AI fits

Once data is collected and stored, it serves more than curves: detecting drift before a breakdown, forecasting consumption, correlating weather and operation. We do not promise it at the start; we first build clean collection, without which no AI is worth anything. When the time comes, integration happens on your real data.

What it costs

A communication driver on a real device takes a few days and removes most of the uncertainty. A complete platform, collection, history, application, alerts, represents a few weeks of engineering, then a monthly operations fee. We always start with the driver: if it does not talk to your equipment, you know before committing to the rest.

References

Proof

What we delivered and operate

  • For a vineyard-cover manufacturer: BACnet/IP driver on a Niagara Edge 10 controller, real-time monitoring platform with a TimescaleDB database, mobile app, alerts and remote commands. Reads in under two seconds against forty-five before.

  • An online quote builder for a metal-cutting company, connected to its manufacturing parameters.

  • A Kubernetes cluster and backup chains operated for SMEs, with monitoring and alerts.

Frequently asked questions

Questions we are asked

Our controllers are old, is it possible?

Most often, yes: Modbus and BACnet have existed for decades and most controllers speak one of the two, sometimes through a gateway. The communication driver, built in a few days on a real device, answers the question before committing to the rest.

Do we have to expose our equipment on the internet?

No, and we advise against it. Collection is done by a service placed on the equipment's network or through an outbound tunnel, never by opening the controller to the outside. Commands go through the platform, authenticated, logged, with per-person rights.

What happens if the 4G link drops?

The equipment keeps working with its own automation, collection resumes and catches up on missing data when the link returns, and the platform flags the missing readings as an alert. It is a case planned from the design stage.

Can we control, or only watch?

Both, if the equipment allows it and if you want it. We always respect local safeties, which keep priority over a remote command, and every command is traced. Some clients prefer to start with monitoring alone and add control later.

How long for a first site?

A few days for the communication driver, a few weeks for a usable platform on a first site. Extending to the next sites is quick, the platform is designed multi-site and multi-organisation from the start.

Which piece of equipment would you like to see from your phone?

Thirty minutes to talk protocol, connectivity and a first pilot site.

Book a free diagnostic

Nous utilisons des cookies pour analyser le trafic et améliorer votre expérience. En savoir plus