The Cost of Changeover in High-Mix SMT
Changeover — the time between the last board of one job and the first good board of the next — is the single largest source of lost productivity in high-mix SMT production. While a placement machine may run at 99% uptime during production, changeover events can consume 20-40% of the total available shift time in factories that switch products 4-8 times per day.
A typical unoptimized changeover takes 30-60 minutes. Multiply that by 6 changeovers per shift, 2 shifts per day, and 250 working days per year, and the math is sobering: 1,500-3,000 hours of line time lost annually to changeovers on a single line. At $200-500/hour in lost production value, that represents $300,000-$1,500,000 per line per year.
The good news: with systematic optimization, 70% of that changeover time can be eliminated. This article provides a practical roadmap based on proven methods used in leading SMT factories.
SMED Principles Applied to SMT
Single-Minute Exchange of Die (SMED) is a lean manufacturing methodology developed by Shigeo Shingo for reducing setup times. The core insight applies directly to SMT: separate activities that must happen while the line is stopped (internal setup) from activities that can happen while the line is still running (external setup). Then convert as many internal activities to external as possible.
Internal Activities (Line Must Be Stopped)
- Swapping feeder trolleys or carts
- Loading the new production program
- Adjusting conveyor width for new board dimensions
- Changing stencil on the screen printer
- Running the first-board inspection
External Activities (Can Happen While Line Runs)
- Preparing material kits for the next job
- Loading reels onto offline feeder trolleys
- Verifying component-to-feeder assignments
- Retrieving materials from storage
- Preparing the screen printer stencil
- Programming or transferring the next job’s recipe
In an unoptimized factory, most of these external activities happen after the current job finishes — making them effectively internal. The first step to reducing changeover is moving every possible activity to external.
The 5-Step Roadmap to 70% Reduction
Step 1: Pre-Kit Materials Before Changeover Starts
Impact: 25-35% of total changeover time eliminated
Material preparation is the single largest changeover time component in most factories. Operators waiting for the warehouse to deliver components, searching for missing reels, or discovering that a needed reel is expired — these events turn a 10-minute feeder swap into a 40-minute ordeal.
What to do:
- Generate the pick list for the next job as soon as it appears on the production schedule
- Begin kitting while the current job is still running — at least 30 minutes before expected completion
- Verify every component against the BOM before delivering the kit to the line
- Stage the completed kit at the line-side before the current job finishes
With automated storage: systems like the Neotel SMD BOX can automatically begin retrieving next-job materials based on MES schedule data. The kit is ready at the output port in 5-10 minutes with zero manual picking or walking.
Step 2: Offline Feeder Setup
Impact: 15-20% of total changeover time eliminated
Loading reels onto feeders while the machine is stopped is one of the most common internal-to-external conversion opportunities. With offline feeder setup, operators load reels onto spare feeders or feeder trolleys while the current job runs.
What to do:
- Invest in spare feeder trolleys — at minimum, one spare trolley per line for the most frequently changed positions
- Set up a dedicated offline feeder preparation area adjacent to each line
- Load and verify feeders for the next job during the current job’s production run
- When changeover starts, swap entire trolleys rather than individual feeders
Requirements: spare feeder trolleys ($5,000-15,000 each depending on size), offline preparation area (2-4 square meters per line), feeder verification scanner.
Step 3: Material Verification Automation
Impact: 10-15% of total changeover time eliminated
Manual component verification — checking that the right reel is on the right feeder in the right slot — is time-consuming and error-prone. Barcode-based verification systems automate this process.
What to do:
- Implement barcode scanning at feeder loading: operator scans reel → scans feeder slot → system confirms match
- Use the placement machine’s built-in verification if available (most modern machines support this)
- Reject mismatches immediately before the line starts — catching errors during setup is 100x cheaper than catching them at AOI
When combined with automated storage: the storage system pre-verifies every reel at retrieval. The feeder-loading scan becomes a second-layer confirmation rather than the primary verification, reducing per-reel scan time.
Step 4: Smart Storage Integration
Impact: 10-15% of total changeover time eliminated
Connecting the storage system directly to the production schedule creates a pull-based material flow. Instead of operators pushing material requests to the warehouse, the system anticipates demand and stages materials proactively.
What to do:
- Integrate the storage system with MES or production scheduling software
- Configure automatic pre-staging: the system begins retrieving next-job materials based on the schedule, not operator requests
- Enable automatic return processing: when a job completes, the system prompts operators to return unused materials and processes them immediately
- Use FEFO logic for MSD components to eliminate floor life checking during changeover
Result: materials arrive at the line before the operator asks for them. The storage system becomes invisible — it just works.
Step 5: Standardized Changeover Procedures
Impact: 10-15% of total changeover time eliminated
Even with the best materials and equipment, changeover efficiency depends on consistent execution. Standardized procedures ensure that every operator, on every shift, follows the optimal sequence.
What to do:
- Document the changeover sequence step by step, including parallel activities (what operator A does while operator B does something else)
- Define clear roles: who handles the printer, who swaps feeders, who verifies
- Time each step and set targets — make changeover time a visible metric on the shop floor
- Conduct regular changeover audits: observe actual changeovers and identify deviations from the standard
- Share best practices across shifts — often the night shift has developed tricks that the day shift does not know about
Real-World Improvement Data
Here is what typical before/after metrics look like when all five steps are implemented:
| Metric | Before Optimization | After Optimization | Improvement |
|---|---|---|---|
| Average changeover time | 45 minutes | 12 minutes | 73% reduction |
| Material wait time during changeover | 15-20 minutes | 0-2 minutes | ~90% reduction |
| Feeder loading time | 15 minutes (at machine) | 3 minutes (trolley swap) | 80% reduction |
| Verification time | 8 minutes (manual check) | 2 minutes (scan-based) | 75% reduction |
| Changeovers per shift | 4 (limited by changeover time) | 8 (changeover is no longer the bottleneck) | 2x throughput flexibility |
| Wrong-component incidents | 1-2 per month | 0 | 100% elimination |
The 70% Reduction Roadmap: Phased Implementation
Phase 1: Quick Wins (Weeks 1-4)
Focus on organizational changes that require minimal investment:
- Implement pre-kitting (start kitting one job ahead)
- Document and standardize the changeover sequence
- Assign clear roles for each changeover step
- Start measuring changeover time per event
Expected improvement: 20-30%
Investment: minimal (labor and training only)
Phase 2: Equipment Enablers (Weeks 4-12)
Add the hardware and systems that enable offline preparation:
- Purchase spare feeder trolleys for offline setup
- Set up offline feeder preparation areas
- Deploy barcode verification at feeder loading
- Establish line-side material staging areas
Expected improvement: additional 20-25% (cumulative 40-55%)
Investment: $20,000-60,000 per line
Phase 3: Automation (Weeks 8-20)
Deploy intelligent storage and integrate with production systems:
- Install automated storage system
- Integrate with MES for schedule-driven material staging
- Configure automatic pre-kitting and FEFO compliance
- Implement automatic return processing
Expected improvement: additional 15-20% (cumulative 55-70%)
Investment: $80,000-250,000 per storage unit
Phase 4: Continuous Improvement (Ongoing)
Optimize based on data:
- Analyze changeover time data to identify remaining bottlenecks
- Optimize production schedule sequence to minimize changeover scope (group similar BOMs)
- Implement predictive material staging based on schedule look-ahead
- Cross-train operators for multi-role changeover capability
Expected improvement: additional 5-10% (cumulative 65-75%)
Quick Wins vs. Long-Term Investments
| Action | Investment | Time to Implement | Impact |
|---|---|---|---|
| Standardize changeover procedure | Low (training) | 1-2 weeks | 10-15% |
| Pre-kitting (manual) | Low (process change) | 1-2 weeks | 15-20% |
| Spare feeder trolleys | Medium ($5-15K each) | 2-4 weeks | 15-20% |
| Barcode verification | Medium ($3-10K per line) | 2-4 weeks | 10-15% |
| Automated storage + MES integration | High ($80-250K) | 8-16 weeks | 15-20% |
| Schedule optimization software | Medium ($20-50K) | 4-8 weeks | 5-10% |
Key Takeaways
- Changeover time is the single largest productivity drain in high-mix SMT — reducing it by 70% is achievable with systematic effort
- Apply SMED principles: separate internal from external activities, then convert internal to external
- Material preparation is the biggest opportunity — pre-kitting and automated storage together eliminate material-related changeover delays
- Offline feeder setup converts machine-stopped loading time to machine-running preparation time
- Standardized procedures and measurement create the discipline for sustained improvement
- Start with quick wins (pre-kitting, standard procedures) and progress to automation for the full 70% reduction
Frequently Asked Questions
- How much can SMT changeover time realistically be reduced?
- Real-world SMT operations have achieved 60–73% changeover time reductions by combining pre-kitting, offline feeder setup, spare feeder trolleys, and intelligent automated storage. A typical high-mix line running 45-minute changeovers can reach 12–18 minutes with a structured SMED-based approach. The full 70% reduction requires addressing all three phases: material preparation, feeder loading, and machine setup — not just one.
- What is SMED in SMT manufacturing?
- SMED (Single-Minute Exchange of Die) is a lean manufacturing methodology originally developed for stamping presses that applies directly to SMT changeover. The core principle is separating “internal” activities (done while the machine is stopped) from “external” activities (done while the machine runs), then systematically converting internal activities to external. In SMT, the biggest SMED opportunity is moving feeder loading and material preparation offline before the current job ends.
- What is pre-kitting in SMT and how does it reduce changeover time?
- Pre-kitting means assembling all the component reels required for the next job before the current job finishes, so materials are ready at the line the moment the changeover begins. It eliminates the material search and retrieval phase from the changeover window — typically 15–25 minutes of a conventional changeover. With intelligent automated storage, pre-kitting is triggered automatically from the production schedule and materials arrive at the output port in feeder-slot order.
- How long does it take to implement SMT changeover improvements?
- Process changes (standardized procedures, pre-kitting discipline) can be implemented in 1–2 weeks with minimal investment. Spare feeder trolleys take 2–4 weeks to procure and deploy. Barcode verification systems deploy in 2–4 weeks. Automated storage with MES integration is the longest investment at 8–16 weeks, but delivers the largest sustained reduction. A phased approach over 3–6 months achieves 70% reduction while spreading capital expenditure.
