Issue link: https://iconnect007.uberflip.com/i/373246
32 SMT Magazine • September 2014 the trUe ImpACt OF AUtOmAtION continues The vendor said that the line would perform as expected if it had been maintained correctly. Maintenance, whether driven simply by time, the number of units produced, or by more so- phisticated calculation of work-load analysis or SPC trend analysis, is clearly a major issue relat- ing to the automation of production. The third lesson that should have been learned is related to data preparation for the automated processes. In the case of SMT, the machines obviously follow their programs. These programs are generated based on the de- sign of the PCB, plus details about materials used, with regard to shape and other physical attributes. The real pain in the process, however, is the many different formats into which the data needs to be converted and subsequently main- tained for each of the many machine platforms potentially in a line setup or across a shop-floor. Inconsistency of performance caused by differ- ences in material shape definitions is common, which results in significant line down-time whenever a new product is introduced or an ex- isting product moved. As more automation is introduced, the lev- el of data preparation and conversion across different formats increases, extending to the three-dimensional space of assembly. Human operators have quite advanced built-in dexter- ity when it comes to assembly, compared with robots that need to have every movement de- fined precisely. Automation for the Future Coming forward to 2014, we see that there is one final lesson to learn from the other three. We now have the overhead of specialised engi- neers in automation. Some are assigned to the optimisation of the automated line, ensuring that the line performs at maximum efficiency yet is flexible for the changes needed. Others are assigned to the analysis of performance in terms of quality and maintenance. Yet more prepare the precise process data that generates the instructions for the automation to execute. The decision to purchase automation is usu- ally based on the replacement of operator head- count, especially in higher cost labour areas in the United States and Western Europe. Seldom is the cost justified on the basis of repeatability and quality alone. How about the cost though of the additional teams of specialist engineers? The cost of an engineer is a great deal more than production-line operators. Our final lesson then is to understand the true cost of ownership of automation. Simple automation, which may be perceived as being cost-effective with an attractive ROI, may hide a lot of costs. More complex automation may represent significantly higher investment, but may be far more cost-effective, even within the medium term. One example of newer, more advanced au- tomation technologies is the Baxter production robot. This example, perhaps looking a little immature to some, represents a far more realis- tic direction for automation. Robots need to be adaptable, have more sensors, degrees of free- dom, and sophisticated software so that they can eliminate the tedious detail of information required for them to operate. In turn, this also brings more ability for the production operators to program the robots with simple instructions. This leads to flexibility in adapting robots to many forms of assembly, thus reducing the need for extensive work at changeover. As with the higher-end SMT machines today, the in- spection for quality can be engineered into the automation itself. I like the direction that industrial robots such as the Baxter example represents, other than the moving eyes on the screen which re- ally creep me out. The future then for automation is certain; production of this kind of advanced industrial robot is on its way. How soon it will be until we see mass adoption? I am not sure, and who will be involved to set up, maintain, and operate the robots again needs to be seen. For now though, keep it coming; we are not quite there yet, but I want it! Smt feATure Michael ford is senior marketing development manager with Mentor Graphics.