6.1 Maintenance Governance: KPIs, Roles, Escalation

Total Productive Maintenance (TPM) is the operational methodology used to maximize the operational efficiency of manufacturing equipment. It shifts the paradigm from "repairing when broken" to "maintaining to prevent failure." In the context of high-volume electronics manufacturing, TPM is the primary driver for minimizing unscheduled downtime and stabilizing process capability (Cpk). This chapter defines the calculation of Overall Equipment Effectiveness (OEE) and mandates the execution of Autonomous Maintenance.

Overall Equipment Effectiveness (OEE)

OEE is the definitive metric for equipment productivity. It is calculated as the product of Availability, Performance, and Quality. The facility target for critical SMT assets is >85%.

Formula: OEE = Availability x Performance x Quality

1. Availability (A)

Measures the percentage of scheduled time that the equipment is capable of operating.

Calculation: (Planned Production Time - Unplanned Downtime) / Planned Production Time
Loss Drivers: Equipment failure, material shortages, setup/changeover time.
Mandate: Setup times (SMED) must be <15 minutes per line changeover.

2. Performance (P)

Measures the speed at which the equipment runs relative to its designed cycle time.

Calculation: (Total Count / Run Time) / Ideal Cycle Time
Loss Drivers: Minor stops (jams, pickup errors), speed throttling (running machines slower than rated spec).
Mandate: Machines must not be throttled below 95% of rated speed without written engineering justification.

3. Quality (Q)

Measures the yield of Good Units produced relative to the Total Units started.

Calculation: (Total Units - Defective Units) / Total Units
Loss Drivers: Solder defects, misplacements, scrap.
Mandate: First Pass Yield (FPY) at AOI must exceed 98.5%.

The Pillars of TPM

Implementation relies on specific pillars tailored to the facility infrastructure.

Pillar 1: Autonomous Maintenance (AM)

Operators are the first line of defense. AM transfers routine maintenance tasks from technicians to machine operators to ensure early detection of abnormalities.

Clean & Inspect: Operators must clean sensors, nozzles, and transport rails at the start of every shift.
Lubrication: Operators must verify lubricant levels and apply grease to accessible linear guides weekly.
Tagging: Operators must apply "fuguai" (abnormality) tags to any loose bolts, leaks, or noises for technician review.

Pillar 2: Planned Maintenance (PM)

Time-based or condition-based maintenance performed by skilled technicians to restore equipment to original condition.

Scheduling: PM tasks are triggered by run-hours (e.g., 1000 hours) or cycle counts, not calendar days.
Parts Replacement: Wear parts (filters, belts, vacuum cups) are replaced preemptively based on Mean Time Between Failure (MTBF) data, not upon failure.

Pillar 3: Focused Improvement (Kobetsu Kaizen)

Cross-functional teams analyze major losses to eliminate root causes.

Trigger: Any downtime event exceeding 60 minutes triggers a mandatory Root Cause Analysis (RCA).
Output: Corrective actions must include hardware modifications or procedure updates to prevent recurrence (Poka-Yoke).

Digital Tracking & Reporting

Manual logs are insufficient for OEE analysis.

MES Integration: All pick-and-place machines and reflow ovens must be networked to the Manufacturing Execution System (MES).
State Codes: Equipment must automatically report state codes (Running, Idle, Error, Setup). Generic "Idle" codes are prohibited; operators must select specific reason codes (e.g., "Waiting for Parts").
Dashboarding: Real-time OEE data must be visualized on line-side Andon boards.

Final Checklist

Metric / Pillar	Parameter	Target / Mandate
OEE	Overall Target	>85%
Availability	Changeover Time	<15 Minutes
Performance	Speed Rating	>95% Rated Speed
Quality	First Pass Yield	>98.5%
AM	Operator Task	Clean & Inspect Daily
PM	Trigger	Run-Hours / Cycles
Improvement	RCA Trigger	>60 Min Downtime
Reporting	Data Source	Auto-MES (No Manual)