Downtime Pareto analysis
Paste the cause and duration columns of your downtime log, one line per stop. See which few causes make up most of the total downtime on a Pareto chart with a cumulative line.
One stop per line: "cause;duration". Separator ; , or tab (pasting from Excel works). If you use a decimal comma, use ; as the separator. Durations of the same cause on several lines are added. A header line is skipped.
The calculation runs in your browser; the log you paste is not sent anywhere.
13 lines used.
4 causes (40% of the 10 causes) make up 82% of the total (threshold 80%).
- Vital few causes
- 4 / 10
- Share of the vital few
- 82.1%
- Share of all causes
- 40%
- Total
- 1,090 min
- Largest cause
- Bearing failure (35.3%)
Pareto chart
- vital few
- Cause
- Cumulative %
- Threshold 80%
Causes
| Cause | Duration | Share | Cumulative |
|---|---|---|---|
| Bearing failurevital few | 385 | 35.3% | 35.3% |
| Changeover / adjustmentvital few | 250 | 22.9% | 58.3% |
| Waiting for materialvital few | 165 | 15.1% | 73.4% |
| Sensor failurevital few | 95 | 8.7% | 82.1% |
| Operator error | 60 | 5.5% | 87.6% |
| Lack of lubrication | 45 | 4.1% | 91.7% |
| Power outage | 30 | 2.8% | 94.5% |
| Belt break | 25 | 2.3% | 96.8% |
| Software fault | 20 | 1.8% | 98.6% |
| Cleaning / jam | 15 | 1.4% | 100% |
share = duration / total · cumulative = running sum of shares · vital few = causes up to the first point where the cumulative share reaches the threshold
A Pareto analysis ranks the relative size of causes; it does not tell you the root cause. The result depends on record quality: inconsistent cause codes or a large "other" bar can make the ranking misleading.
Let's look at which of the leading causes could be spotted in advance with sensor data.
Request a conversation01
How to use it
A
Paste your downtime log as "cause;duration" lines (copying two columns from Excel is enough) or load the example data.
B
Choose the threshold (80% by default) and read the bars and the cumulative line on the Pareto chart.
C
Focus on the causes marked "vital few"; copy the result into your report.
02
The Pareto principle and downtime data
The Pareto principle rests on the observation that most results come from a few causes. In downtime data this usually shows up as a handful of cause codes making up most of the total downtime. This tool ranks causes by duration from largest to smallest and computes each cause's share and the cumulative share.
"80/20" is a rule of thumb, not a law: your data may give 70/10 or 80/40. What matters is how steeply the curve starts. A steep start means focusing on a few causes pays off; a nearly flat one means downtime is spread over a wide range and a single action will not be enough.
03
How to read the chart
Bars show the duration of each cause (left axis), the line shows the cumulative percentage (right axis). The dashed horizontal line is your threshold; the bars up to the point where the line first crosses it are the "vital few". The chart shows at most 12 bars; the remaining causes are grouped into one "Other" bar, while the table lists all of them.
If many causes are grouped under "other" and that bar is large, your cause codes may be too scattered or vague. Simplifying the classification first makes the analysis more meaningful.
04
Duration, frequency and cost: what to rank by?
Ranking the same data by duration, by number of stops and by cost can give different results. A rare but long failure ranks high by duration; frequent short stops rank high by count. This tool ranks by the value you enter; if you enter cost per stop instead of duration you get a Pareto by cost.
A Pareto is a starting point: it does not say why the largest cause occurs. Root cause analysis (for example 5 Whys) and sensor data complete the step.
FAQ
- What is a Pareto analysis and what is it for with downtime data?
- A method that ranks causes by impact from largest to smallest and shows the cumulative share. With downtime data it shows which few causes to direct limited maintenance resources at to cut total downtime the most.
- Is the 80% threshold mandatory?
- No. 80% is a common habit; you can set the threshold anywhere from 60 to 95. The threshold only decides where the "vital few" end; the ranking and shares do not change.
- Should I enter number of stops or duration?
- Whichever matters to you. Duration shows lost production, count shows the intervention load, cost combines both. Running all three and comparing them is informative.
- Can I write the same cause on several lines?
- Yes. Durations on lines with the same cause name are added; case and leading or trailing spaces are ignored. Spelling differences (for example "Bearing failure" and "Bearing fail") count as separate causes, so code your records consistently.
- How do I paste from Excel?
- Select the two columns (cause, duration), copy and paste into the box; Excel separates them with a tab, which the tool recognises directly. A header line is skipped automatically.
Find the cause of downtime in your data
Let's look at which of the causes that cost you the most time could have been spotted in advance with machine data, and what that would do for downtime.