Issue link: https://iconnect007.uberflip.com/i/1010078
AUGUST 2018 I SMT007 MAGAZINE 43 Changeover down-time between setups of different products are taken as a fact of life. Loss of machine program efficiency to support the grouping of materials between products to reduce change-over time, is an often-ignored consequence. These "hidden" additional costs, then appear on top of the line-rate issues, which incidentally have also now worsened, due to the increase in the number of different products being produced, each of which now needs their own separate buffer stock. Factories were always going to lose this battle against the customer's supply-chain, as stock held in the supply-chain is far more visible and is under the control of the customer in the EMS case, or by the sales organization in the OEM case. Legacy Planning Solutions To try to make sense of this whole situa- tion, there are literally hundreds of commer- cial solutions available for factory planning. Earliest attempts at shop-floor planning solu- tions came from ERP, which were spectacu- larly poor in dealing with the complexities of PCB manufacturing machines, processes and steps. It is not possible for a simple time- per-product model definition to be an accu- rate representation. Consequently, as a next step, more complex, process simulation tools became popular, to analyze the performance of the machines and processes, in a connected line and across lines, to identify a more accu- rate timing model for products at each produc- tion configuration, which could then create better schedules, squeezing out more produc- tion performance. In performing fixed simula- tions of production requirement, these solu- tions completely missed the point that the most important losses were based on the grow- ing customer demand fluctuation. Next came advanced planning and scheduling (APS) tools that focused on more complex schedul- ing requirements that came with higher mix production. Attempting to create a long-range, complex schedule became little more than a complete failure, since as soon as customer requirements changed, the whole remainder of the plan was obsolete. Another critical issue is that the APS software is trying to work with its hands tied behind its back, as without digi- tal remastering of the product data, the legacy product to line configuration dedication gave very little flexibility, and no opportunity for APS algorithms to find alternatives. The resul- tant value of using APS tools, even in the best of cases, were significantly less than expected. We then come to the most practical and established shop-floor planning solution, currently in use in the majority of electronics manufacturing, Excel. Sometimes simple, sometimes extraordinarily complex, Excel remains the practical tool of choice for real-world work-order scheduling. It is an intensively manual process to maintain, as expected performance data needs to be continuously updated, as does the progress of manufacturing as time passes. In the hands of the most expert, experienced and expensive shop-floor planning engineers, Excel shines over and above other legacy tools in keeping the factory running. Due to this ability of Excel to assist and support human decision- making in planning in the very near term, demand fluctuations become manageable, not easy, but possible, which is more than can be said for the other legacy solutions discussed. The losses incurred by legacy solutions on a routine basis may be as little as 20% of lost productivity, measured in real terms of factory and machine availability, but may also often be as high as 80% in higher mix and lower volume combination situations. The worst thing about this is that often, these losses are not reported, and nothing is done to eliminate them. Instead, it is thought that high degrees of flexibility, together with high productivity are Sometimes simple, sometimes extraordinarily complex, Excel remains the practical tool of choice for real-world work-order scheduling.