Overall Equipment Effectiveness?OEE tells you, in one number, how well the machine ran during the current 5-second window.
OEE = Availability × Performance × Quality
85%+ is world-class; 60% is industry average. Look at the three sub-metrics to see which lever moved.
A 0% reading means at least one of the three was 0 — usually Quality during a fault recovery window.
--%
Availability — · Performance — · Quality —
Availability?Availability answers: was the machine running when it was supposed to?
Avail = RUN seconds / total seconds
Penalised by breakdowns, jams, e-stops, changeovers. Planned breaks don't count if you flag them as scheduled downtime.
--%
— RUN / — down
Performance?Performance answers: when it ran, did it run as fast as it should?
Perf = avg BPM ÷ ideal BPM
Below 100% means micro-stops or slow cycles. Ideal BPM is the nameplate speed (28 832 here).
The dashed green line on the chart marks the ideal BPM target.
--%
Avg — / Ideal — BPM
Quality?Quality answers: of what we made, how much passed QC?
Quality = (bottles − rejects) / bottles
Bottles ticked by the counter; rejects by the reject sensor.
Computed per 5-second window. If the machine was faulted (no bottles) but the reject counter still ticked, you'll see 0%. Read it alongside Availability.
--%
— rejects of —
Bottles Produced?Bottles Produced is a session-wide running total — the sum of every per-window delta on bottle_count_total since the dashboard opened.
Refreshing the page resets the counter. Kafka still holds the full history.
0
Cumulative this session
Reject Events?Reject Events is the running session total of failed bottles flagged by the reject sensor.
The smaller number below counts alarm episodes — contiguous spans where any machine alarm was active.
Reject % = rejects ÷ bottles
An alarm episode often produces many rejects in quick succession.
0
0 alarm episodes
Production Throughput · BPM
LIVE · 5s WINDOW · MEAN