A line we commissioned last year logs about four thousand tags. On a good week somebody looks at nine of them. That is not a failure of the system, it is what the system is for.
01Collection is the easy part
It has been the easy part for a decade. Gateways are cheap, OPC UA works, storage costs nothing worth discussing. Any competent integrator can have your whole plant historised in a fortnight, and plenty of projects stop there and declare success, because the dashboard looks impressive in the review meeting.
02The dashboard problem
A screen with thirty numbers on it is a screen nobody reads. It gets glanced at, it becomes wallpaper, and within a month the only person who opens it is the person who built it. The instinct to put everything on one screen because it is all available is the single most reliable way to make sure none of it is used.
03Start from the decision
Ask what someone will do differently. A shift lead can act on: are we at rate right now, what stopped us most this shift, and is anything about to stop us. Three questions, three answers, one screen, readable from across the hall. Everything else is for analysis later, and analysis later is a different tool for a different person.
04Alarms are the same problem, louder
An operator who receives four hundred alarms a shift acknowledges them without reading them. That is not carelessness, it is the only possible response. Alarm rationalisation is dull work and it is usually the highest return activity available on an old line: every alarm has to have a cause, a consequence and an action, or it is not an alarm, it is an event.
05The test
Six months after handover, go and look at what people have open. If it is the screen you built, it works. If it is a spreadsheet somebody made themselves, they have told you exactly what they needed and it is not what was delivered. That spreadsheet is the most valuable document in the plant and you should go and read it.



