Do we need sensors on the equipment?

For optimal prediction, sensors are ideal (vibration, temperature, current). However, Modulario also works with manually entered operational values or data imported from existing SCADA systems.

Which types of equipment does the system monitor?

Production lines, compressors, pumps, HVAC units, CNC machines, elevators, forklifts, boilers, refrigeration systems — essentially any equipment with measurable operating parameters.

How does the system know what is a normal vs. abnormal reading for a specific piece of equipment?

During the first 4–8 weeks the system collects data and establishes a baseline of normal operation for each asset. After that it begins detecting deviations from that baseline.

AI Predictive Maintenance: End of Unplanned Downtime

An unplanned production line stoppage costs on average 5× more than planned maintenance. An emergency service call-out is 2–4× more expensive than a scheduled one. And that excludes lost production, a missed customer delivery deadline, or safety risk.

Reactive vs. Predictive Maintenance

Reactive (the common state):

Equipment breaks down
Unplanned production pause
Urgent spare parts order (express surcharge)
Emergency service — 2–3× higher price
Repair: 2–5 days

Preventive (calendar-based):

Maintenance every X days/hours regardless of actual condition
Parts replaced that could still have run longer
Unnecessary labour costs

Predictive (with AI):

System monitors equipment parameters continuously
Identifies anomalies → predicts failure 7–21 days ahead
Schedule maintenance when the production plan allows it
Order exactly the parts needed
Service technician arrives prepared with the right components

What AI Monitors

Equipment Parameters (from sensors or SCADA):

Vibration (abnormal vibration = bearing about to fail)
Temperature (overheating = cooling or friction problem)
Electrical current (increased draw = mechanical resistance)
Noise and ultrasound (cavitation, wear)
Hydraulic pressure

Operational History:

Running hours since last maintenance
Number of starts/stops
Records from previous service interventions

Output: Predictive Service Plan

Every week the service manager receives a summary:

PREDICTIVE SERVICE PLAN — Week 19

⚠️ CRITICAL (within 7 days):
- CNC Mazak VCN-430 — spindle bearing
  Failure probability: 78 %
  Recommended action: Replace bearing, order SKF 6205-2RS

🟡 SCHEDULED (7–21 days):
- Compressor Atlas Copco GA55 — filter
  Failure probability: 52 %
  Recommended action: Replace air filter

✅ OK: 12 assets

ROI: Calculation for a Manufacturing Company

One unplanned production line stoppage (6 hours):

Lost production: 3,600 €
Emergency service: 800 €
Express spare parts: 400 €
Total loss: 4,800 €

Planned maintenance (same scope):

Scheduled service: 300 €
Spare parts without express: 250 €
Production pause within planned window: 0 € extra loss
Total cost: 550 €

Saving from one averted breakdown: 4,250 €

AI Predictive Maintenance — Fix It Before It Breaks