Every plant has a number on a whiteboard for how much the line ran last week. Almost nobody has an honest number for how much it didn’t run, and why. Unplanned downtime is the quietest expense on the floor: it does not show up as a line item, it shows up as overtime on Friday, a late shipment, and a shift that felt busy and shipped nothing. If you have read our plain-English take on OEE, downtime is the “Availability” question — was the machine even running when it was supposed to be? This page is how you get that number down without buying anything.
What actually counts as unplanned downtime
Unplanned downtime is any time a line is scheduled to run and isn’t, for a reason you didn’t choose. A breakdown is the obvious one. But the ones that add up are smaller and quieter: waiting for material, waiting for a forklift, waiting for QC to sign off, a jam that takes four minutes to clear eleven times a shift, an operator who stepped away because the next job’s paperwork wasn’t ready. None of those feel like “downtime” in the moment. Add them up across a week and they usually dwarf the one big breakdown everybody remembers.
Planned downtime — a scheduled changeover, a maintenance window you chose — is a different thing, and it is not the enemy. If your real problem is changeovers eating the day, that is its own fix, and we wrote it up separately: how to reduce changeover time.
Why the real number is always bigger than the one on the board
Most floors only log downtime when it is long enough to be embarrassing — the twenty-minute stop, the hour waiting on a part. The three-minute stops never get written down, because writing them down takes longer than the stop itself. So the board says the line was down two hours and the truth is closer to six, spread across sixty small stops nobody logged. That gap is exactly where the money is, because small repeated stops are the cheapest kind to fix. They are almost never a broken machine. They are a process that makes the operator stop.
Start with the log, not the fix
The instinct is to jump to a solution — a new sensor, a maintenance contract, a faster machine. Don’t, yet. For one week, have each line keep a dead-simple downtime log: when it stopped, roughly how long, and one phrase for why. Paper on a clipboard is fine. You are not building a database; you are finding the top three reasons, because on almost every floor three reasons cause most of the lost time. You cannot fix sixty causes. You can fix three. This is the same “follow the time, not the opinion” approach behind our walk through where your plant floor is losing time.
The cheapest fixes almost always come first
Once you have the top three, most of them turn out to cost nothing but attention. Material arriving late to the line is a staging problem, not a machine problem — fix where and when material shows up and the stops disappear. A jam that recurs in the same spot is usually a guide, a worn part, or a misfeed that a five-dollar fix ends permanently. A stop for “waiting on QC” is a scheduling problem — the check is happening at the wrong moment. The point of starting with the cheap fixes is not that they are small. It is that they are fast, they build trust on the floor, and they clear the noise so the one genuinely expensive problem is finally visible on its own.
When the bottleneck is the only machine that matters
Here is the part most downtime programs get wrong: not all downtime is equal. An hour lost on a machine that already has material piling up in front of it costs you nothing — the line was going to wait on the next station anyway. An hour lost on your single slowest station costs you an hour of everything you could have shipped. So before you spend a dime reducing downtime, know which machine is the one that sets your pace. That is your bottleneck, and downtime there is the only downtime that reaches the loading dock. We wrote a full guide on how to find the bottleneck — do that first, then aim your downtime effort at it.
Maintenance you schedule beats maintenance that ambushes you
Once the quiet stops are handled and you know your bottleneck, the last lever is the big one: the breakdown you didn’t see coming. The fix is not a bigger budget, it is a calendar. Breakdowns are rarely random. The bearing that seized had been getting hot for weeks; the belt that snapped had been fraying. A short, boring, scheduled check — on the machine that matters, on a rhythm you choose — turns a four-hour surprise on the worst possible day into a twenty-minute swap on a slow Tuesday. You do not need a software system to start. You need a list of the two or three machines whose failure stops everything, and a standing time to look at them before they look at you.
What “good” looks like
A well-run line is not one that never stops. It is one where the stops are short, known, and mostly chosen. You will always have downtime; the goal is to move it out of the “ambush” column and into the “scheduled” column, and to shrink the pile of tiny stops nobody was tracking. When you get there, the Friday overtime stops being automatic, the ship dates stop being a gamble, and the whiteboard number and the real number finally start to agree.
Want the three reasons your line stops?
That is exactly what the first visit is for. The free plant-floor walkthrough is a few hours, on your floor, watching where the time actually goes — no software to buy, no system to rip out, ERP-agnostic, family-owned. You get the top reasons your line stops and a plain plan to fix them. If nothing else, you will finally have the honest number.
