Autonomous Maintenance Step One: Clean and Inspect

The grease build-up on the side of machine seven had been there for eleven months. Nobody had reported it. Nobody had cleaned it. It had become part of the scenery.

That is how most factories run. Not through negligence, exactly, but through familiarity. When something has always looked a certain way, it stops registering as a problem. The fault hides in plain sight until the bearing seizes or the seal fails and the line goes down at the worst possible moment.

Autonomous maintenance exists to break that pattern. And it starts with the simplest act there is: cleaning the machine properly and paying attention to what you find.

What Autonomous Maintenance Actually Means

Most maintenance regimes work like this: operators run the machine, maintainers fix it when it breaks. The two groups barely speak. The operator notices something odd, files a ticket three days later, and by then the bearing has failed and the line has been down for six hours.

Autonomous maintenance turns that around. It gives operators ownership of their equipment, starting with the most basic act: cleaning it and looking at what they find.

This is not about cutting the maintenance headcount. It is about putting informed eyes on the machine every single day. Eyes that belong to someone who knows what normal looks and sounds like.

Why Cleaning Is Not Just Cleaning

There is a phrase used in TPM circles: cleaning is inspection. It sounds like a slogan, but it describes something real.

When an operator wipes down a machine with a cloth, they cover every surface. They notice the oil seep on the gearbox that was not there last week. They feel the vibration in the guard that has started to rattle. They see the discolouration around the seal that suggests heat is escaping somewhere it should not.

A maintenance technician doing a scheduled PM visit will check the items on their list. The operator doing a daily clean will find the thing that is not on any list yet.

The hidden defect, the one that causes the unplanned stoppage three weeks from now, is usually visible well in advance. It just takes someone to look.

How to Run the First Clean-and-Inspect

The first time is the hardest. Machines that have not been cleaned in years have layers. There is contamination on contamination, and finding the baseline takes effort.

Block out time for it properly. One hour per machine is not unusual for the initial clean. You are not rushing this. You are establishing what normal looks like so that future deviations stand out.

Go section by section. Clean each part of the machine, then inspect it before moving on. Use a simple checklist, nothing fancy, just a prompt for what to look at: lubrication points, seals, belts, guards, fasteners, electrical cabinets, cooling vents. The checklist is not the point. The looking is the point.

Tag everything that is not right. Red tags for defects the operator cannot fix themselves. Blue tags for things they can address with basic cleaning or tightening. Do not try to fix everything on the day. Record it, report it, then work through the list systematically over the following weeks.

Photograph the defects as you find them. A photo takes three seconds and gives the maintenance team a precise location and a baseline to compare against at the next inspection.

What You Will Find

In a factory that has not run a clean-and-inspect programme before, the first round will typically surface between eight and twenty defects per machine. Most of them will be minor. Some will not be.

Common finds include lubrication lines that have cracked or disconnected, meaning a bearing has been running dry. Coolant leaks dripping onto electrical components. Guards that were removed at some point and never replaced. Contamination in places that affect product quality without the operator realising it.

None of these things are catastrophic on their own. Together, they represent the baseline condition the machine has been running in. That condition is where your unplanned downtime comes from.

Turning the Find Into a Habit

The value is not in the first clean. It is in the second, the fifth, the thirtieth, when the operator notices a difference from what they have seen before.

Build the clean-and-inspect into the shift start routine. Fifteen minutes. No exceptions. The consistency matters more than the thoroughness of any individual session. A quick look every day beats a thorough look every quarter.

Over time, operators develop a feel for their equipment that no PM schedule can replicate. They start to predict failures before the machine tells them. That is the goal.

If you are working out where autonomous maintenance fits into a broader improvement programme, the six-step framework on the /six/ page sets out how clean-and-inspect connects to the later stages of operator skill development and equipment ownership. For teams that are still mapping where their losses actually come from before committing to a full AM rollout, the loss analysis tool at losses.cadencestandard.com is a practical starting point. It takes less than twenty minutes and gives you a clear picture of whether contamination and minor stoppages are your real problem, or whether something else is driving the numbers.