ANDON·WALL
Plant Throughput OEE
● EXAMPLE · demo data, not live
SHIFT A · DAYS · Plant Aurora-3
--:--:--
← Operations
Plant OEE · Live ▲ 1.1 vs YDA
87.2%
Overall Equip. Effectiveness
on pace · world-class band
ref ring ≥85% · target 85%
94.1%
Availability
96.0%
Performance
96.5%
Quality
Line A
318u/hr
OEE 89%go
Line B
324u/hr
OEE 90%go
Line C ◂ constraint
286u/hr
OEE 82%spindle vib ↑
Line D
331u/hr
OEE 91%op free
Throughput
1,240u/hr
vs 1,200 tgt · 103%go
Shift Output
9,640
/ 9,400 target · 102%
Quality
99.1% FPY
Scrap 0.7% · under 1%go
Reliability
18min DT
MTBF 142h · MTTR 0.4h
Sched Attain
98%
1 SKU changeover behind
Shift-Output Forecast · AI
9,720
projected final · +3.4% vs 9,400
burn-up slope · ETA target 15:00on pace
Predictive Maintenance
Line C spindle vibration +14% over baseline · failure probability 31% in next 8h. Recommend inspection at next changeover 11:40 — prevents the only amber on the board and protects 6% recoverable throughput.
Bottleneck finder · Line C constraint, rebalance 1 op Anomaly baselining · cycle-time + vibration envelopes Auto cause-code · pre-fills stoppage reason Scrap-spike predictor · EoS projection vs 1% line Shift-output forecast · 9,720 / 9,400
Downtime Pareto · This Shift18 min total
Changeover
7 m
Material wait
5 m
Jam
4 m
Maintenance
2 m
▸ tap to open the timestamped event log →
Shift Burn-Up · Actual vs Target+2% above line
Actual · 9,640 to 13:42 Target pace · 9,400 on pace 9,720 by 15:00
Plant Vitals
Schedule Attainment98%
Units this shift9,640
Unplanned downtime18 min
First Pass Yield99.1%
Operators on floor14 / 16
Field Guide
Read the wall like the floor
A 20-foot-legible briefing on how a plant manager runs the shift off this screen.

How to use this

  • 1Andon dot first, number second. Scan the four line tiles — green means hitting target, amber is your constraint. Today Line C is the only amber: that is where the shift is being lost, full stop.
  • 2The hero gauge is your "go / no-go." Plant OEE ≥85% (world-class ring) means the floor is healthy; the A·P·Q satellites tell you whether a dip is uptime, speed, or quality before you walk the line.
  • 3Read the burn-up slope, not the snapshot. If the lime actual line sits above the dashed target pace, you finish over plan — the slope is the early-warning, the total is the result.
  • 4Let AI prevent the red. The predictive-maintenance strip is the highest-value pixel: catching the Line C spindle at the 11:40 changeover turns an unplanned stoppage into zero lost minutes.
  • 5Rebalance, don't add labor. The bottleneck finder pulls one operator from your best line (D) to the constraint (C) for 6% recoverable throughput — free capacity, no overtime.
For your own floor: if you had to choose ONE number for a wall a worker sees 200 times a shift — would it be OEE, units/hr, or andon color? Most plants pick andon color, then regret hiding the constraint behind a single roll-up.

Watch the walkthrough

Four AI agents walk this dashboard.

In context

Sample feed
Illustrative — wire to your MES / historian + sensor feed.
Plant OEE percentile
vs sector benchmark (discrete mfg)
P82
▲ top-fifth
Energy / unit
kWh per finished unit, shift avg
1.84
▼ 4.1%
Spindle vibration · Line C
mm/s RMS vs 4.5 baseline
5.1
▲ +14%
Cycle-time CpK · constraint
process capability, target ≥1.33
1.41
▲ 0.06
Days since recordable
plant safety streak
214
— hold