A single line, watched continuously over four months. Planned downtime fell from 33% to 8% and output rose 11%, but the more valuable result was what the data kept surfacing next: the emerging faults and quality drifts a one-off audit would have missed.
Plenty of sites will run a one-off study, get a number, act on it, and move on. This line was left under continuous monitoring for four months instead, and the difference is the whole point. A snapshot tells you where you stand today. A trend tells you where you are drifting, and it keeps telling you long after the first round of fixes is done.
The obvious wins came first. The line's planned downtime dropped from 33% to 8% once the data showed how much of that "planned" time was habit rather than requirement: changeovers that ran long, buffers built in for problems that no longer happened, stoppages nobody had revisited. Recovering that time helped lift output by 11% over the period.
Unplanned stops get attention because they hurt. Planned downtime is treated as fixed and rarely challenged, which is exactly why it was hiding the largest, easiest block of recoverable time.
The more interesting story was what the data caught week to week. A cycle time creeping up on one station. A reject rate ticking higher on a particular product. A minor stop cause appearing more and more often. None of these were visible on any single day, and none would have shown up in a one-off audit. On a four-month trend they stood out clearly, and each was addressed as an emerging issue rather than after it had already cost a shift.
That is the case for leaving monitoring running rather than sampling: every fix resets the baseline, and the next problem is already on the chart before it becomes a stoppage. Improvement stops being a project and becomes a routine.
Planned downtime is treated as fixed, so it rarely gets questioned. Measuring it showed that much of the 33% was habit rather than necessity, and most of it could be recovered.
Slow drifts. A creeping cycle time, a quality reject rate ticking up, a stop cause appearing more often week on week. These only show up when you compare this week to the last three months.
No. Because the data kept running, each improvement had a new baseline, and the next emerging issue was already visible on the trend before it cost a shift.
Book a demo with one of our manufacturing experts.