← Raymond P. Capisinio

ENTRY 10 · 2026·07·22 · Shift Log

Close the Loop, Not the Log

The best part of learning to build software isn't the software. It's what it shows you — how much a plant could have, and usually doesn't.

The best part of learning to build software isn't the software. It's what it shows you — how much a plant could have, and usually doesn't.

On the floor, a fix waits for the next shutdown. In the software that tracks it, a better idea can be live this shift — if you're willing to change the shape.

Almost any plant knows the bad week: twenty issues piled on one machine. The formal channel exists — maintenance tickets raised against each item — but they live in legacy software, behind an interface that fights the person raising them. So the real story scatters into phone calls, old emails, and permits — and there's still no straight answer to "what was actually done about this?"

For years, the structure was locked — not by choice, but by that same software. It was built once and couldn't adapt, so every ticket hung on the equipment tree. That was the only shape the system knew.

Here's the strange part: mineral processing is not a slow industry. We put expert systems on grinding circuits in the 1990s and cameras on flotation froth two decades ago. When intelligence comes packaged from a vendor, we buy it. But the everyday layer — logs, tickets, handovers — could only be bought, never bent. So it froze.

Then the constraint quietly expired. AI has been writing working software for about five years now — and for the last three, the language you build in is plain English. Not code, not a computer-science degree: the same words you'd use to brief a new technician. So the people who know the plant best no longer need a software company — they already speak the only language it takes.

Most plants haven't noticed, and every year of not noticing is another year of operating history captured badly or not at all. The exact data an AI-ready plant will wish it had.

We lagged behind for what?

The move here wasn't a new feature. It was a new shape — the same tickets, re-hung so the loop closes where the work actually happens.

That's the builder's advantage: when you can reshape the tool, a structure everyone treated as fixed becomes an afternoon's change.

What in your plant is everyone treating as a law, when it's really just a habit?

Get the next piece.

Get the free field guide — The Handover Gap: three disasters, one lesson, and a checklist for your next shift. Plus a short note when a new piece goes up. No noise, no pitch.

Run a shift somewhere? I compare notes with operators and supervisors on how real floors handle handover and capture — what works, what quietly doesn't. No pitch; I'm not selling anything. Tell me how yours does it: [email protected]