Predictive Maintenance Fundamentals

# Predictive Maintenance Fundamentals

Manufacturing maintenance has evolved through four maturity levels: reactive (fix it when it breaks), preventive (service on a schedule), predictive (service based on condition), and prescriptive (AI recommends optimal actions). Most factories are stuck between reactive and preventive. AI makes the jump to predictive practical and affordable.

The Maintenance Maturity Spectrum

Reactive — Fix it when it breaks. Lowest planned cost, highest total cost due to emergency repairs, production losses, and cascading failures. Typical unplanned downtime: 15-25% of production hours.

Preventive — Service on a calendar or run-time schedule. Reduces unplanned failures but creates unnecessary maintenance (replacing parts that have remaining life). Typical improvement: reduces unplanned downtime to 5-12%.

Predictive — Monitor equipment condition and service when data indicates degradation. Extends part life, reduces both unplanned failures and unnecessary maintenance. Typical improvement: reduces unplanned downtime to 2-5%.