Predictive Model OutputModel: Bearing Degradation Model v2.3 (synthetic)
PREDICTIVE MAINTENANCE
Question answered: “WHAT IS LIKELY TO HAPPEN?”
Model outputs
Prediction PM-2026-184 · 22 Sep 2026, 14:04
M-103 CNC Machining Center
Predicted failure
Spindle Bearing Degradation
Failure probability
78%
Prediction horizon
5–8 days
Confidence
86%
Remaining useful life
≈ 142 op. hours
Contributing signals
Vibration increase +35%38%
Spindle RMS velocity 5.8 → 7.8 mm/s over 7 days
Temperature increase +18%24%
Housing temperature 74 → 87°C
Historical similarity 84%22%
Matches 11 of 13 past bearing-wear signatures
Operating load High16%
Sustained 18.4 h/day at 88% spindle load
Feature contribution to predicted probability (relative share).
Failure probability — last 7 days
Prediction explanation: The combination of increasing spindle vibration, temperature deviation and sustained operating load resembles historical degradation patterns associated with bearing wear.
Prediction Output
“Inspection or preventive maintenance is recommended within 5 days.”
But the prediction cannot answer…
But WHEN should maintenance occur?
Which asset should be prioritized?
What is the production impact?
Are technicians and parts available?
Not visible to the predictive model: Production schedule Customer orders & SLAs Spare-part inventory Technician rosters Backup capacity Cost of downtime
CONTINUE TO DECISION INTELLIGENCE