HyperVision MES.
A shop floor that reports itself.

The module the system learns from when the plant is supposed to be working. You enter the shift roster, breaks, days off, planned inspections and operation time standards. Without it every minute of the day counts the same and the system cannot tell a planned stop from a breakdown — and that difference is exactly what decides whether availability in OEE means anything.

Losses become visible against the time the machine was genuinely meant to be producing.

The black box of the process. It records signals from machines and utilities without a break — temperatures, pressures, currents, energy use — with a timestamp. When a defective batch or a breakdown shows up, you go back to a specific hour and see how the process behaved.

The cause is reconstructed from the record, not from the memory of whoever was on shift.

Documentation shown where it is needed: on the screen by the machine, in the context of the order being run. The operator sees the instruction, the drawing or a short video for that specific part. The system enforces versions — a withdrawn instruction disappears from the stations immediately, with nobody walking the floor.

A new person gets into the order faster, and a product changeover stops being a moment of risk.

A register for the events nobody wants to keep in a notebook: accidents and near misses, customer complaints, quality non-conformities. Every report has an owner, a deadline and a status, and closing one requires describing the corrective action.

You walk into an audit (IATF, ISO, GxP) with a complete history instead of a binder.

Control points placed in the middle of the process. You define what has to be checked between operations and how often — dimension, weight, appearance — and the system enforces the frequency and holds the batch until the result is confirmed.

A defect surfaces at the operation that created it, not at packing, when the part already carries its full cost.

A separate record of what went into the bin. The operator reports the scrap and picks a reason from a code list, so after a month you have material losses broken down by cause. The same reports feed the quality component of OEE.

You can see which causes account for most of the loss — and which are worth touching at all.

Rules that push an event to the right person when they are not looking at a screen. You decide what is worth interrupting someone for — a stoppage longer than a set time, a critical machine failure, a parameter out of range — and who should receive it. Channel of your choice: SMS, e-mail, WhatsApp, Telegram or another messenger the plant already uses. If nobody responds within the set time, the report escalates a level up.

The information reaches the person who can act, without hunting for them across the floor.

The place where a call from the floor turns into a ticket with a history: who picked it up, how long the repair took, what was replaced. Planned service lives in the same module — periodic inspections, replacements by running hours, calibrations — so breakdowns and the inspection plan are visible next to the production schedule.

Maintenance works from one list of events instead of phone calls and notes.

The data layer underneath reporting. Events from the floor are cleaned, arranged into a model and exposed in a form ready for Power BI or another reporting tool. This is the piece that makes the system on the floor and the board report count from the same definition.

One source of numbers for production, quality and the board.

Data exchange with the systems already running in the company, over an API in JSON. Most often with the ERP: orders and item codes come down to the floor, while output, consumption and scrap go back up — with nothing retyped and no file circulating by e-mail. Scope and format are agreed during the integration audit, because every system exposes data differently.

Data is entered once, in one place.

A predictive layer over the data the other modules collect. The system learns a machine’s normal behaviour and signals when the pattern starts to drift — rising vibration, a lengthening cycle, creeping temperature — before it ends in a stoppage. It builds on the history from the Machine operating history module, so it delivers its full value once data has accumulated, not on day one.

Some interventions can be moved from breakdown to planned downtime.

Guidance that walks the operator through a procedure that has to run the same way every time: line start-up, changeover, restart after a breakdown, cleaning, first-article validation. Every step has a description and a pass condition — until the result matches, the procedure does not move on. What remains at the end is a record: who went through the whole sequence, when, and with what result.

Instead of a signature on a sheet, what remains is evidence that the procedure was completed in full.