SMT007 Magazine


Issue link:

Contents of this Issue


Page 19 of 93

20 SMT Magazine • March 2016 before changing products. Machine program ef- ficiency and line balancing were the critical ele- ments for high-volume manufacturing because a change of as little as 1–3% in line performance would be reflected almost directly on the bot- tom line of the operation. When changing a line from one product to another on an SMT line, the key bottleneck was the removal of all materials of the out-going product, followed by the setup and verification of all of the materials for the incoming product. This could typically take hours, which was not such a problem for the high-volume environ- ment because it only happened every few days and offered the opportunity to perform essen- tial and routine maintenance on the SMT ma- chines. The real issue came as the changeovers on dedicated lines became more and more fre- quent, perhaps once every two days, and then once every day. The time that it took to per- form the changes to the materials was now much more significant, as was the sheer volume of materials that would be set up for the more frequent line changeovers, as well as the flood of materials that now needed to be returned to the warehouse. The immediate solution came from SMT pro- gramming engineers who realized that differ- ent products could be made on dedicated lines by keeping sets of materials in place that were common to all products. If the products within the group were variants of each other and had an almost identical bill of materials, product changeover between products within the group could happen without any loss time at change- over from material setup. Even products with roughly similar bills of materials could be in- cluded in the common material setup grouping, perhaps with just a few materials needing to be exchanged at changeover if the total number of material feeders needed by the whole group of products exceeded the capacity of the machines in the line. This approach succeeded at reducing the changeover time of SMT lines, but at a price. The program efficiency of SMT machines is highly dependent on the physical setup of the materi- als with respect to the number of placements of each material and in which area of the PCB they of the time; but this is clearly not the case, es- pecially when visiting any factory doing high- mix production. When calculating the absolute productivity, or asset utilization, the figures are actually closer to 40%. No matter what effort a factory makes, the absolute asset utilization of SMT machines and other related complex processes declines sharp- ly as the number of product changes increases. This decline is caused by the extensive setup time when changing between different mod- els. With the trend of increasing product mix and decreasing lot sizes, the problem is growing worse…and it's not going away. The fluctuation of customer demand is brought more directly to the factory, as the stock holding in the distribu- tion chain is reduced to save significant costs to the overall business. If we are to introduce fur- ther automation into PCB assembly, we must be able to achieve the return of investment goal that it should be capable of while making an automated factory that is flexible and efficient. And to do this, we need to once and for all re- solve the issue of decreasing productivity. Simply adding more automation is not a panacea These goals will greatly affect the choice of new automation and how it is implemented. The SMT machine is a classic example of this "no-win" situation with automation and flex- ibility. A line of high-performance SMT ma- chines dedicated to making a single product used to be the normal state of manufacturing, where dedicated lines would run for many days thE buyEr'S guIdE to autoMatIon " with the trend of increasing product mix and decreasing lot sizes, the problem is growing worse…and it's not going away. "

Articles in this issue

Archives of this issue

view archives of SMT007 Magazine - SMT-Mar2016