Use case

Live OEE for every critical asset

Machines, lines and critical assets publish their state, cycle counts and rejects to ConnectMQ. Your OEE dashboard, MES and APM tools subscribe and show availability, performance and quality per asset, line and site — as it happens, not in tomorrow’s report.

No card required

Example message

PUBLISH acme/dallas/packaging/filler-02/counts

{"ts":"2026-10-22T09:00:00Z","state":"running","good":11842,"reject":96,"ideal_cycle_s":0.3,"shift":"B"}

QoS 1 · TLS 8883 · PUBACK

The problem

OEE numbers that arrive a shift too late

In many plants OEE is still put together from operator sheets, MES exports and spreadsheets at the end of the shift. Each site counts downtime differently, comparing plants takes weeks, and by the time a performance loss shows up, the shift that could have fixed it has gone home.

How it works

  1. 1

    Capture state and counts

    A gateway reads PLC tags — machine state, good and reject counts, ideal cycle time and downtime reason codes — over OPC UA, Modbus or S7.

  2. 2

    Publish per asset

    Every change is published to the asset’s topic. Current state is retained, so any dashboard knows at once whether a machine is running, idle or down.

  3. 3

    Calculate OEE

    Your OEE engine, MES or APM tool subscribes with one wildcard per site and calculates availability, performance and quality per asset, shift and product.

  4. 4

    Compare and act

    Dashboards, andon boards and the ConnectMQ live monitor show the same numbers, and threshold rules alert the shift leader when an asset has been down too long.

ConnectMQ

What it needs from the broker

Retained current state

Retained state topics give every new screen, MES instance or report the current state of each asset without waiting for the next change.

One topic tree across sites

The ISA-95 workspace keeps enterprise, site, area, line and asset in one tree, so headquarters compares plants with a single subscription.

Many consumers, one source

OEE dashboards, MES, historians and the 3D Digital Twin all subscribe to the same messages — no extra load on the PLC for each tool.

Access per site and team

Separate credentials and per-topic ACLs: each plant publishes only to its own topics, while corporate teams read across all sites.

What you get

  • One definition of availability, performance and quality across plants
  • Stops and slow cycles visible during the shift, not after it
  • Downtime reasons captured at the source, next to the counts
  • New lines and sites added by publishing into the same tree
Example topics
acme/dallas/packaging/filler-02/state      (retained)
acme/dallas/packaging/filler-02/counts
acme/dallas/packaging/filler-02/downtime
acme/dallas/packaging/filler-02/oee        (retained)

Industries that run this use case

Frequently asked questions

How do I calculate OEE from MQTT data?

Publish machine state changes, good and reject counts and the ideal cycle time per asset. Your OEE application subscribes and calculates availability (run time vs planned time), performance (actual vs ideal output) and quality (good vs total). ConnectMQ delivers the data; the calculation runs in your OEE tool, MES or dashboard.

Can I track asset performance across multiple plants with one MQTT broker?

Yes. Each plant publishes into the same ISA-95 topic structure — enterprise/site/area/line/asset — with its own credentials. A central dashboard subscribes with a wildcard and sees every site, while each site team can be limited to its own topics.

Does this work for assets outside the factory, like pumps or compressors?

Yes, the same pattern applies. A cellular or site gateway publishes run hours, state and performance values for each asset. Persistent sessions and QoS 1 help when links are intermittent, and Last Will shows when a site goes offline.

Build it on ConnectMQ

Start on the Free plan — no card required — connect your first devices today and upgrade when you go live.