On a coupled line, only one machine actually stopped.
A jam at the palletiser stops the labeller, which starves the filler. Three machines record a stop and one of them caused it. VZIAM MES keeps that difference: each stop is timed by its own controller and qualified as the machine’s own or induced by another, so the work goes to where the cause is.
Everything on the line reports downtime, and none of it says which.
High-speed packaging fails in ways that are individually trivial and collectively expensive.
Cascades counted three times
When a stop propagates, every machine on the line logs one. The Pareto then ranks the machines that suffered rather than the one that caused it, and maintenance is sent to the wrong end of the line, repeatedly, because the ranking keeps saying the same thing.
Short stops below the threshold
A thirty-second jam cleared by the operator standing next to it is not usually declared. Across a shift on a fast line those clearances add up, and they are a category of loss that nobody has a figure for.
Changeovers timed from memory
Format changes are recorded, if at all, as a duration somebody estimated afterwards. Whether one team is genuinely quicker than another is therefore a matter of opinion, and it gets settled the way opinions do.
Time every machine, then say which stops were its own.
The order the machines stopped in is a fact the controllers already hold. Each step is a part of VZIAM MES, and its page is linked under it.
- 01
Time each machine from its own controller
Every machine on the line has its own timeline, with no gaps and no overlaps, timed from its controller’s state changes. Which machine stopped first is then a matter of timestamps, not of who reached the radio first.
How machine data is read - 02
Qualify a stop as own or induced
Each stop reason is typed under AFNOR NF E60-182, and induced is one of the types: a filler starved by the labeller is waiting, not broken. Because the type travels with the reason, the losses a machine caused are kept apart from the ones it suffered.
Telling cause from consequence on a coupled line - 03
Count the jams nobody declares
An unexplained stop under the threshold set for that machine is counted as a micro-stop, against performance, without asking anyone for a reason. The count is complete, which it never is when it depends on someone choosing to log it.
Micro-stops, counted - 04
Time format changes from the machine
A format change is a stop timed by the controller, qualified with its reason and compared with the target set for it. The cockpit shows the overrun as it happens, and comparing one changeover with another needs nobody on the line with a clipboard.
How a stop is timed and qualified
Four people, one account of the line.
The operator, maintenance, the supervisor and the production manager read the same stops, each with the machine that had it and the type its reason gave it.
The machine times the stop, the operator says why.
Each machine’s timeline comes from its own controller. The operator is asked only for the reason, from the list set for that machine, and an induced reason says the stop came from elsewhere.
How a stop is timed and qualified- Start and end
- From the controller
- Reason
- From the operator
- Type
- Own or induced
Sent to the machine that caused the loss.
Stops typed as induced are kept apart from the machine’s own, and micro-stops stay out of the breakdown Pareto and MTBF. What reaches maintenance is the machine’s own failures, not the ones it suffered.
Maintenance inside VZIAM MES- Own stops
- Per machine
- Induced stops
- Kept apart
- Micro-stops
- Out of MTBF
The short jams, counted every one.
An unexplained stop under the threshold set for that machine is a micro-stop: counted from the controller, booked against performance, with no reason asked of anyone. After 15 seconds the cockpit already shows the line hesitating.
Micro-stops, counted- Threshold
- Set per machine
- Pre-alert
- After 15 seconds
- Booked to
- Performance
The line’s state, above the line.
Any browser on a smart TV shows each line’s state, OEE and cadence against nominal, readable from across the hall, with alerts on the line’s card as they are raised.
Andon and TV displays- State
- Per line
- Cadence
- Against nominal
- Alerts
- With a type and severity
What gets connected on a packaging line.
- Each machine
- Its own timeline
- Controllers
- OPC UA, MQTT, HTTP push
- Older machines
- Modbus, S7, via a gateway
- Stop types
- Closure, planned, induced, own, micro
- Micro-stop threshold
- Per machine
- Changeover
- Planned or not, with a target
- Line display
- Any browser on a TV
Frequently asked questions.
Tell us which machine your team thinks is the constraint, and ask the engineers how the record would confirm it.
Book a callNo. It times every machine from its own controller, so the order in which they stopped is on the record, and the operator qualifies each stop with a reason typed own or induced. The cause comes from those two facts together, not from an inference the software makes on its own.
Every stop the controller reports is timed. An unexplained stop at or under the threshold set for that machine is counted as a micro-stop; a longer one asks the operator for a reason. Both delays are settings, per machine.
That depends on how the reason is typed. A changeover typed as planned reduces the opening time rather than availability; typed as unplanned, it is an availability loss. Either way its duration comes from the machine, and its overrun against target is shown.
Yes. Each machine’s signals are mapped once (an OPC UA node, an MQTT topic or an HTTP post), so each has its own timeline from its own controller rather than one state for the whole line.
Yes. Any browser on a smart TV or a small box shows each line’s state, OEE and cadence against nominal, and alerts appear on the line’s card as they are raised, each with a type and a severity.
Other industries we work in.
Plan the change before it reaches the line.
Request a demo and an engineer looks at how your production runs today, shows how the software would fit around it, and where a pilot on one line would start.
01Pick a time
Choose a slot in the calendar, and note what you make and what reads your machines today.
02Screen shared with an engineer
A video call with someone who builds the software, in English or French.
03What would connect
For each machine, its path: native (OPC UA, MQTT or HTTP push), through a gateway, or nothing to build.