The right trained operator at every critical station — even in mask and goggles.
A camera at each critical stage recognises the operator and checks that they are trained and allowed there. It logs who worked where and for how long — primary, alternate, someone else or an unknown face — and how long a stage stood empty, with daily and monthly reports.
Brazing and other critical stages are done by whoever is standing there; the record says nothing about who it was or whether they were trained.
“A camera at each critical stage recognises the operator — mask and goggles on — checks they are trained and allowed, and logs who worked where and for how long.”
Drawn, not promised: the top row is how plants answer "who was at the brazing stage?" today — from the roster, at audit time. The bottom row is the camera recognising the operator, mask and goggles on, and checking the skill matrix as the shift runs.
Hover or tap a stage to see what is captured there.
Not a standalone tool — a product on the MOS platform, sharing one login, one set of people and roles, and one database with every other module. The lit layers are the ones this product uses; the signal path shows how data moves through them.
The honest map, drawn as one thread. Filled nodes are in the product today; outlined nodes are other MOS modules on the same platform; dotted nodes are roadmap and never represented as available until built. Hover a node.
Each card ends with the thing itself as it appears in MOS — not a description of it.
Step through it or let it play. Each step: who does it, on what, what is captured, and the check that guards it.
Operators are enrolled with photos and the stages they are trained for; primary and alternate per stage.
An IP camera the plant already has covers the critical stage; the GPU machine runs recognition.
The person at the stage is recognised, even in mask and goggles.
Is this person trained and allowed at this stage?
Who worked at the stage and for how long; empty-stage time is counted.
An untrained person at a critical stage, or a stage left empty too long, raises an alert with the picture.
Daily and monthly presence per stage and shift.
| Your ERP / Tally typically gives | Operator Verification with AI adds |
|---|---|
| Whoever was free ran the stage | The person at the stage recognised and checked |
| No record of who did the critical work | Presence per stage, per shift, with duration |
| Found out after the field failure | Untrained-at-station alert at the time |
| Badges that get swapped | A face, in mask and goggles |
IP cameras at each critical stage — usually the ones already there
One GPU machine on the plant network
MOS Edge in the plant
No badge, no wearable
What it connects to, where it runs and what it gives you — the short version your IT and your plant head both want.
Enrolment, recognition in PPE, presence log, untrained alerts, empty-stage time, reports.
MOS Skill Matrix supplies who is certified for what.
Camera position and lighting are set per stage at installation.
Attendance from stage presence — not represented as available.
Capabilities depend on the plant's process and what is connected at site. Nothing on this page is represented as available until it is built and running.
Who is where, right now
Proof the right person did each critical job
Presence by stage, day and month
Untrained-person exceptions
An IP camera facing where the operator stands.
A GPU machine on the plant network that does the recognition.
A few photos of each operator, taken once.
Critical stages chosen; cameras aimed and tested.
Every operator enrolled with the stages they are trained for.
Exceptions, empty-stage time and monthly reports running.
Yes — it was built on brazing stages where operators wear both.
On the plant's own GPU machine and server, inside the plant network.
Every visitor registered, approved by the host, and a QR gate pass — with a live list of who is inside.
See the module →Every cycle checked against the right way of doing the job.
See the module →Who can run which station — and where the line is exposed.
See the module →Tell us about your plant and what you want to see. We visit, connect one machine or station, and you watch your own data in MOS before anything is quoted.