CNC Downtime Tracking

Downtime is rarely one big breakdown. It is dozens of short stops and long waits that never make it onto a log. Here is how to measure them, put a price on them, and get the hours back.

By Ryan Lenhart · Updated Oct 5, 2026

Ask a shop floor how much downtime they had yesterday and you will hear about the alarm on machine 4 and the broken tool on the lathe. Ask the machines and you get a different answer: a dozen gaps between cycles, a 40-minute wait for a first-article sign-off, a program that was not ready at shift change, and an hour where the operator was running two other machines.

That second answer is the one worth acting on. CNC downtime tracking means capturing every stop during planned production, classifying the big ones, and reviewing them often enough to fix the causes that repeat.

The three kinds of stop — keep them separate

When these are lumped into one "downtime" number, maintenance gets blamed for scheduling problems and nobody owns the waiting. Separating them is the single most useful thing a downtime system does.

The causes behind most CNC downtime

From the timelines of typical job shops, the recurring losses look like this:

1. Waiting on material or fixtures — the job is released but the stock or fixture is not at the machine.
2. Waiting on programs or prove-outs — the next job's program is not posted, or needs edits at the control.
3. Setups and changeovers — especially the parts of setup that could happen while the previous job is still running.
4. First-article and in-process inspection — the machine sits while the part walks to the inspection room and back.
5. Operator coverage — one person running three machines means two of them wait.
6. Tool problems — tools not preset, not in the crib, or breaking early.
7. Alarms and faults — real, but usually fewer minutes than all the waits combined.
8. Shift change and breaks — machines that stop for 20–40 minutes around every shift boundary.

None of these show up as one big event. They show up as minutes, many times a day, on many machines.

Why manual downtime logs miss most of it

Paper logs and end-of-shift spreadsheets capture the stops people remember. In practice, short stops — under about 10–15 minutes — are routinely missed or rounded away, and stop times get written from memory hours later. The result is a log that undercounts downtime and over-represents the dramatic events.

Manual input still matters for one thing: why. The machine knows it stopped. The operator knows it was waiting for the inspector. The best systems combine automatic timing with a light human reason code.

How to track CNC downtime — step by step

Step 1: Define planned production time

Write down which machines are expected to run on which shifts. Downtime only means something against a plan.

Step 2: Capture machine state automatically

Use monitoring software that reads running, idle, stopped, and alarm states from the machine. FEAI Machine Monitor does this through MTConnect on supported machines and builds a minute-by-minute timeline.

Step 3: Set a short reason code list

Start with something like:

• Waiting — material
• Waiting — program
• Waiting — inspection
• Waiting — operator
• Setup / changeover
• Tooling
• Alarm / maintenance
• Planned stop

Add a code only when it would change a decision.

Step 4: Code the big stops, not every stop

Ask operators to explain stops longer than a threshold — say 10 minutes. The timeline already has the rest.

Step 5: Review daily, act weekly

A ten-minute morning review of yesterday's longest stops, with one owner per stop, does more than a monthly downtime report nobody reads.

Step 6: Verify the fix

After a change, compare the same machine on the same shifts. If the stop category did not shrink, the fix did not work.

How to reduce CNC downtime once you can see it

Start with the bottleneck machine. An hour saved there is an hour for the whole shop. An hour saved on a machine that waits for work anyway changes nothing. See Find CNC Bottlenecks for how to identify it.

Downtime and OEE

Downtime is the availability part of OEE. If a machine planned for 480 minutes stops for 96, availability is 80% before speed or quality losses are counted. Reducing downtime is usually the fastest way to raise OEE in a job shop. Try the numbers in the OEE calculator.

Tracking downtime with FEAI Machine Monitor

FEAI Machine Monitor records machine state through MTConnect on supported CNC machines and gives you:

• A minute-by-minute timeline of running, idle, stopped, alarm, and offline time
• Stops ranked by total duration and frequency, by machine and shift
• Alerts when a machine stops or stays idle longer than you expect
• Utilization, availability, and OEE from the same data
• Exportable reports for the morning meeting
• FEAI Intelligence (Growth and above) that highlights recurring downtime and setup problems across machines

Start tracking downtime today

Connect your bottleneck machine and one more, free, and look at a full week of real stops. Start free — 2 machines.