Issue link: https://iconnect007.uberflip.com/i/674146
68 SMT Magazine • May 2016 not for back-end final assembly. As such, they often cannot be used without costly modifica- tion. Entry/exit openings may be too small to fit a pallet sitting on a conveyor belt with a large assembly in a nest on top. The distance between the tool and the product inside the machine may be too short to accommodate a larger final assembly. A camera might be mounted too low and its lens may need to be replaced with a wid- er one. As an example, we recently had to order a custom modification to a standard gel printer and the custom modification cost us more than the original machine. Standard components such as ICs and most passive components are often packaged for easy removal from tubes, trays, ammo packs or reels. However, there are no common packaging stan- dards for odd-form or customized components. Likewise, packaging for finished goods could be difficult to fully automate in a cost-effective manner. Plastic or cardboard boxes are bulky and difficult to work with, and still need to be brought to and from the line by human mate- rial handlers or by mobile transporters. Boxes need to be sealed and opened. Labels need to be printed, applied and read. Box inserts need to be inserted and removed. Manual or Semi-automated Processes I believe that some processes will remain manual or semi-automated for a long time to come, especially in Asia where labor costs are still significantly lower than in Western facto- ries, despite the constant rise of labor cost. The loading of very small, or very large, or odd-form, components into lead frames, plas- tic frames or other odd-shaped assemblies is an error-prone process that benefits from dexter- ity in handling and immediate visual feedback; at the same time, ideal for a trained operator and challenging to automate. As component and product designs improve and standards de- velop, however, I believe we will see more and more automation in this area. Placing finished goods in their final packag- ing is relatively easy to automate today. How- ever, I believe the last step—adding any inserts, closing the box, labeling it and transporting it—will remain manual processes for some time. Likewise, unpacking direct and indirect mate- rial at the warehouse, transporting them to the production line, and loading the trays, tubes or reels to the loading stations will also remain manual processes for the time being. Finally, visual inspection of products against scratches, dents, discolorations and other de- fects is something that is quite easy (albeit bor- ing) for a human quality inspector to do, but difficult, expensive and time-consuming for an automated inspection system to execute, espe- cially for larger and more complex products. I predict that we will continue to have manual visual inspection steps in our high-reliability manufacturing processes for a long time to come. At Integrated Micro-Electronics Inc. (IMI), we recently developed and installed a semi- automated production line for the assembly of complex automotive ECUs for power steer- ing at our factory in Jiaxing, China (Figure 1). This multimillion-dollar line has more than 25 machines automatically executing complex process steps such as heat staking, dispensing, mounting, resistance welding, visual inspec- tion, electrical testing, screw driving, solder- ing, laser marking and cleaning. The whole line has only three manual operators perform- ing loading and unloading operations, with additional material handlers, quality inspec- tors and equipment technicians to support. We are now working to develop a larger, even faster version of this line at our factory in Gua- dalajara, Mexico. There are always numerous challenges to overcome when implementing such a com- plex line, but we believe that such semi-au- tomated assembly is the future for high-reli- ability, high-quality, high-volume electronics manufacturing. SMT Michael Hansson is the director for global automation at Integrated Micro-electronics Inc. How FaR doES it MakE SEnSE to autoMatE?