Key Takeaways: Changeover time is consistently the largest single OEE availability loss in multi-SKU manufacturing operations, often 15-30% of available production time. SMED (Single Minute Exchange of Die) is the proven methodology for reducing it.
Fabrico provides the data foundation for SMED: accurate changeover baseline measurement, identification of the highest-value changeover types, and post-SMED OEE tracking that verifies whether the improvement held. Operations that use Fabrico-data-driven SMED programs consistently achieve 30-50% changeover time reduction within 90 days.
When manufacturers list their top OEE improvement priorities, changeover time reduction rarely leads the list despite consistently representing 15-30% of available production time in multi-SKU operations. The reason: changeover time feels unavoidable. Product variety requires changeovers. Changeovers take time. The time appears in the OEE availability calculation as a large, chronic, seemingly fixed loss that gets accepted as the cost of serving diverse customer demand.
This acceptance is a mistake. Changeover time is not fixed. In most manufacturing operations, 30-50% of changeover duration is waste, non-value-adding activities that happen during the changeover because they've always been done that way, not because they're genuinely required. SMED methodology, applied systematically to the right changeover types, consistently eliminates this waste, recovering substantial OEE availability without capital investment.
The challenge that prevents most operations from capturing this value: they don't have accurate changeover time data. They know changeovers take "about 3 hours" on Line 5, but they don't know whether that's 2.5 hours on Tuesday mornings and 4.5 hours on Friday afternoons, whether the color changes take longer than material changes, or which specific activities within the changeover consume most of the time.
Without this data granularity, SMED teams identify changeover improvement opportunities through observation and experience, useful but inefficient. With Fabrico OEE data, they identify opportunities through a ranked analysis of actual changeover performance across hundreds of documented changeover events.
Fabrico classifies changeover time as a distinct OEE availability category, separate from unplanned downtime, planned maintenance, and production runs. This classification is configured per line and per changeover type, enabling granular analysis that plant-level OEE numbers can't provide.
The Fabrico changeover data structure captures:
Changeover duration: Machine stop time to machine restart time, measured from PLC signals. Every changeover is timestamped with actual duration, not the planned duration, and not an operator estimate, but the actual time from stop to first good part.
Changeover type: Product-to-product changeovers, size changeovers, color changeovers, and material changeovers each have different root causes, different standard times, and different improvement potential. Fabrico captures changeover type from production order data, when a run for Product A ends and a run for Product B begins, the intervening stop is classified as a Product A to Product B changeover.
Changeover time distribution: For each changeover type, Fabrico tracks the statistical distribution of changeover duration, average, minimum, maximum, and the variance between events. The minimum duration observed represents the "best achievable" performance for that changeover type, what SMED methodology calls the "best of best." The gap between average and minimum is the improvement target.
Time of day and crew patterns: Changeover duration often varies systematically by shift, day of week, or crew assignment. Fabrico surfaces these patterns, if changeovers consistently take 30% longer on the night shift than the day shift for the same changeover type, the cause is operational (training, supervision, or procedure adherence) rather than equipment-related.
With 30 days of Fabrico changeover data, a SMED team has the analysis input that previously required weeks of manual observation: ranked changeover types by total annual time lost, statistical distributions showing improvement potential for each type, and crew/time patterns identifying where standardization gaps exist.
Not all changeovers are worth improving equally. The SMED target prioritization that Fabrico data enables:
Frequency × Duration ranking: The highest-value SMED targets are changeover types that occur most frequently AND take longest. A changeover that happens 3 times per week and averages 4 hours creates 12 hours per week of changeover-related OEE loss. A changeover that happens once per month and averages 6 hours creates 1.5 hours per week. SMED investment should target the 12-hour/week loss first.
Fabrico's AI Agent calculates this ranking automatically, every changeover type is ranked by total annual hours lost, providing an objective priority list for SMED investment that doesn't require manual analysis.
Variance analysis: High variance changeovers, where some events take 2 hours and others take 5 hours for the same changeover type, indicate that a best practice exists but isn't consistently followed. The 2-hour events represent achievable performance; the 5-hour events represent the gap from inconsistent execution. SMED in high-variance situations focuses on standardization rather than fundamental process change, often the fastest improvement path.
Low-variance changeovers where all events consistently take 4 hours indicate that the current process is being executed consistently, but the process itself has improvement potential. SMED here requires process redesign, parallel activity planning, or staged setup changes that can be partially executed before the previous production run ends.
Regression identification: Fabrico monitors changeover time trends over rolling 4-week windows. If changeover duration for a specific type has been increasing over the past 6 weeks, from 2.5 hours to 3.2 hours average, this regression indicates tooling wear, procedure drift, or personnel change that's degrading performance. Identifying and addressing this regression before it becomes the new baseline prevents "normalization of deviation."
The weakness of most SMED programs is the lack of sustained measurement after the improvement event. Teams invest in a SMED kaizen event, reduce changeover time in the workshop, declare success, and move on. Six months later, changeover times have drifted back to pre-improvement levels, because the process change wasn't reinforced, the standard work wasn't maintained, or the original root cause of the long changeover wasn't fully addressed.
Fabrico's post-SMED verification closes this loop automatically. After a SMED improvement event, Fabrico continues monitoring changeover duration for the target changeover type and compares post-event performance to the pre-event baseline.
The comparison shows three things:
A SMED program that reduces average changeover time from 4 hours to 2.5 hours on a line running 3 changeovers per week recovers 4.5 hours of production time per week. Fabrico verifies this: the OEE availability for that line increases by the changeover time reduction as a percentage of available production time. The recovery is visible in the OEE trend and financially quantified by Fabrico's production value calculation.
This verification closes the SMED investment loop: investment in changeover time reduction is justified by OEE data, targeted by changeover analysis data, executed with SMED methodology, and verified by post-event OEE monitoring. Each successful SMED event builds the evidence base for the next improvement program investment, creating a continuous improvement culture grounded in data rather than management intention.
The complete SMED implementation cycle, measurement, target identification, improvement, verification, requires data from both OEE monitoring and maintenance execution. Fabrico provides both in the same platform, eliminating the analysis overhead of cross-referencing two systems.
The specific integration points that matter for SMED programs:
Changeover OEE connected to CMMS tooling history: Fabrico connects changeover duration data to the CMMS records for tooling and mold maintenance.
When changeover times for a specific product transition are increasing over time, Fabrico can correlate this with the maintenance history of the tooling involved, revealing whether the longer changeover is caused by tooling wear that generates quality adjustment time during startup, or by tooling condition that makes the changeover itself take longer.
SMED improvement work orders in CMMS: SMED improvements that require maintenance actions, tooling modifications, fixture changes, connection point improvements, are tracked as CMMS work orders in Fabrico. The work order completion date becomes the reference point for post-SMED OEE verification, allowing Fabrico to automatically compare pre- and post-improvement changeover performance with the correct time boundary.
Interactive Planning Board integration: Fabrico's planning board shows changeover windows alongside maintenance windows. Production planners who schedule changeovers can see which changeover types are currently targeted for SMED improvement and plan accordingly, scheduling SMED-targeted changeovers during pilot periods to maximize data collection and minimize the production impact of the improvement events.
This integration. SMED methodology using Fabrico OEE data for targeting, CMMS for improvement execution, planning board for scheduling, and OEE trend for verification, turns changeover time reduction from a periodic improvement project into a continuous data-driven program.
For manufacturing operations where changeover time represents 15-30% of available production time, this program represents 5-15 OEE percentage points of recoverable availability, the largest single untapped OEE improvement opportunity in most multi-SKU operations.
Login
Blog
