Issue link: https://iconnect007.uberflip.com/i/1545666
38 SMT007 MAGAZINE I JULY 2026 Ideally, everything for the next job is on hand when the machine stops. The only delay occurs for a physical swap between jobs. This change in process may require creating a dedicated prepara- tion cell on the shop floor for feeder loading, mate- rial verification, and cart staging. Based on expert opinion, a preparation cell can shorten the changeover time by about 33% or, in this example, from 105 to 70 minutes. That amounts to saving $1.45 million. Not bad for a three- week analysis. Phase 3: Feeder Cart Standardization Over the next two months, the next steps are to eliminate individual feeder handling. If the opera- tors are still swapping feeders one at a time on an idle machine, that can take 20 minutes (±5 minutes) for a job needing 80 feeders, at an average of 15 seconds per feeder to swap. By setting up a cart with all the feeders ahead of time, the operator can swap the entire feeder cart at the machine in 3–5 minutes. That represents a 14% time savings, or a bit more than $700,000. Astute readers will be pointing out (correctly) that the cart still needs to be loaded with feeders; that is not a zero-time activity now. Process optimization means keeping the line running as close to 100% as possible. While it may be an equivalent activity with respect to labor charges, removing 15 minutes per cart from the changeover time brings significant revenue that can be considered entirely margin: more capacity for the same price. At this point, you're down to 50 minutes per changeover. Standardization Rules Assign Cart A to High-run products and Cart B to Medium-run products. Then use a Universal Cart for Mixed production. Use a slot strategy in which carts store common components in a fixed location. Figure 3 gives an example. Phase 4: Barcode and Digital Verification Our next objective is to spend two months removing setup mistakes. At the completion of this phase, we will have spent approximately four months on this initiative. At the start of this phase, opera- tors are visually inspecting reels, feeders, and slots. This process can be error prone. Familiarity can breed complacency, as well. Once an operator has looked at, touched, and loaded 80 feeders, they are inclined to believe they'd have caught any errors already. What is needed is a single-check procedure in which automation handles verification. This is exactly the sort of function that compute power is good for. With a new scanning system (presumably barcode), operators scan the work order, reel, feeder, and slot. The system confirms the compo- nent is correct, is from the correct lot, loaded into the correct feeder, and in the correct position. Not only does this help reduce redundant checks, but it also reduces setup time. In this case, the estimate is 10 minutes out of the original 105 minutes, about 9%, amounting to about $400,000. After Phase 4, setup times are now in the 40-minute range, resulting in an annual capacity increase of a bit more than $2.5 million. Phase 5: Parallel Setup Teams For process improvement folks, the highly linear nature of traditional setup processes has obvious limitations. Not only should as many things be done externally as possible, but as many things as possible should be done in parallel. This brings us, in Phase 5, to the consideration of elapsed setup time. Traditionally, one operator performs each setup step in sequence: 1. Material 2. Machine 3. Tooling 4. Verification But once these activities have been moved into their own work cell, as they were in Phase 2, Table 3: Example of a slot strategy SLOT COMPONENT F1 10k resistor F2 100nF capacitor F3 1k resistor F4 10uF capacitor