⌘K
ScopePeriod

Diesel Genset GN-01

Asset 360 · Generator · Mill A
GN-01RUNNINGCriticality BGenerator
Manufacturer
Perkasa Diesel
Model
PK-800
Location
Mill A · Utilities
Commissioned
2015-06-27
Age
11 years
Operating hours
64,556 h
Utilization
81%
Availability (90 d)
97%
MTBF / MTTR
2,707 h / 4.3 h
Replacement cost
Rp 2.25 bn
Sensors
temperature, current, oilCondition, rpm
Region
Sumatra
80/100
Health score

Predictive maintenance

Asset Failure Prediction v3.2 · gradient-boosted survival model on IoT, SCADA and CMMS history · refreshed 06:52

Confidence79%
Health score
80/100
Failure probability
12%7 days
Remaining useful life
181days
Anomaly score
0.210–1
Likely cause
No abnormal failure mode detected
Recommendation
Continue condition monitoring; next PM as scheduled
What the model sees
Confidence79%
All monitored signals are within normal bands. Failure probability 12% over 7 days; continue condition monitoring and scheduled PM.
Why this prediction — top drivers (pp of failure probability)
Baseline failure rate for class ≈ 12% · bars push probability up or down
Temperature deviation
+3.4 pp
Age / operating hours
+2.8 pp
Motor current / load
+2.1 pp
Oil condition index
+1.4 pp
Speed deviation
+0.8 pp
Recent preventive maintenance
−4.0 pp
◀ lowers predictionraises ▶

Sensor trends & failure forecast

60-day history (daily) · dashed = model forecast

Temperature · °C
Warning 78 · Alarm 88 °C
No degradation trend — forecast shown only when the model detects a trajectory toward the failure threshold.

Maintenance Decision Intelligence

Weighs failure probability, criticality, downtime impact, repair vs replacement cost, spare availability and production schedule

Decision inputs
Failure probability (7 d)12%
CriticalityClass B
Downtime impactRp 165 m / h
Planned repair costRp 67 m
Replacement costRp 2.25 bn
Spare availabilityFLT-GN-OIL: not in stock — procure
Production schedulePeak crop · next shutdown Sun
Option suitability
Monitor81
Failure risk low; condition monitoring sufficient
Schedule maintenance47
Fits next planned shutdown; balances risk and production
Replace asset23
Health 80/100 and age 11 yrs do not justify capex
Repair now2
Stops the line during peak crop — restan and emergency premium
Replace component2
No single failing component identified
Scenario comparison
OPTION A
Repair immediately
Stop line now, replace component using emergency transfer
Failure risk2%
Expected costRp 483 m
Planned downtime4 h

Lowest risk, but interrupts peak-intake processing and creates restan.

OPTION B
Repair in low-production window
Replace at next planned shutdown
Failure risk4%
Expected costRp 146 m
Planned downtime4 h

Balances failure risk against production loss at the next planned shutdown.

Recommended
OPTION C
Continue monitoring
Run to next scheduled shutdown (14 days)
Failure risk12%
Expected costRp 109 m
Planned downtime0 h

Recommended — risk is low; keep condition monitoring and standard PM.

Recommendation: Monitor
Confidence80%
Continue condition monitoring; next PM as scheduled. Option C expected cost Rp 109 m vs Rp 109 m if left to run; residual risk 12%.
Estimated impact based on synthetic scenario assumptions. Recommendation requires validation by the Maintenance Manager.

Spare parts availability

Key parts for GN-01 across the warehouse network · home store WH-01

Cross-module: Inventory AI
FLT-GN-OIL · FLT-GN-OIL
Lead time 14 d · Rp 0/pcs
0 in home store
Not stocked — procured on demand.
AVR-GN-1 · AVR-GN-1
Lead time 14 d · Rp 0/pcs
0 in home store
Not stocked — procured on demand.
Open Spare Parts Intelligence

Maintenance records

CMMS history, open work orders, failure events and spares

No open work orders — asset on standard PM schedule.
Synthetic asset data and simulated model output. Maintenance recommendations are decision support requiring validation by qualified personnel.
V-TEKI — IT & Business ConsultingDeveloped by V-TEKI · IT & Business ConsultingPrime Agri AI Intelligence Platform — prototype. All operational records are synthetic; model outputs are simulated; financial impacts are scenario estimates; recommendations require human validation. No actual company confidential data is used.