Issue link: https://iconnect007.uberflip.com/i/1464867
54 PCB007 MAGAZINE I APRIL 2022 Automation and computers have been my pas- sion for the last 55 years, both as a student and a professional. This article will introduce the method- ologies that have served me over the years in auto- mating electronics manufacturing. The hardest step is getting started. I know this because I have designed and built portions of nine automated PCB factories in my career, and a dozen more process factories, all automated by comput- ers and the software that drives them. This article focuses on: • Who should design it? • What should they learn? • How do you analyze and plan the automation? • How long should it take? • When is the right time? • How much will it cost? Let's get into some detail on these key questions to help you to get started. Who Should Design It? Independent factory automation consultants are exceedingly rare and, if you find one, they will most likely charge you for learning your processes and procedures. They will be restricted to their products which then restricts their choices of solution. It usu- ally is a better investment to select your own engi- neers who are familiar with your processes and pro- cedures and let them learn how to do the automa- tion planning. Then they can teach others, as the factory will need to pass along this skill repeatedly. One recommended team configuration is having two engineers: a process engineer (usually a chemi- cal engineer) and a control/data engineer (usually an electrical engineer). Having access to a statistician (either a university or a consultant) will be extremely useful. Both engineers should be capable of work- ing independently, with the capability to quickly learn new skills, including statistical tools and data analysis. New college graduates, especially with a graduate degree, would be suitable. A bonus and compensations need to be established, as success- ful factory automation engineers are highly prized and heavily recruited. What They Need to Learn There are six major topics to be addressed with the primary one being to meet your company's business objectives. The company should already have some long-term business objectives that will shape the priorities for an automation strat- egy and product roadmap. Achieving these objec- tives depends upon the other six topics listed here: 1. Zero waste: Waste of any sort, including mate- rials, rework, repair, paperwork, environmental, or workforce, are all suitable focus for benchmarking (or best practices) analysis to improve performance. 2. Internet-of-things sensors: New sensors to provide real-time data are the basis for most Smart factory automation. Sensors can be purchased to provide relevant information and integrated by low- cost programmable logic controllers (PLCs) for stor- age in the cloud. Some may have to be a DIY sen- sor, like the specific gravity unit or amp-hour unit from Chapter 4 of Automation and Advanced Pro- cedures in PCB Fabrication. 3. Predictive analytics: Using new sensor data stored in the cloud, regression analysis, and DOE procedures can create predictive models (digital twins) to reduce or eliminate defects or enhance product specifications like impedance control and registration. 4. Zero-defect manufacturing: Unless your yields are 100%, defects are your largest waste and bench- marking will provide project focus. 5. Driving zero-downtime: Downtime for equip- ment or process affects the bottom line and delivery performance; new sensor data and predictive mod- els can eliminate this. 6. Create solution templates: Documenting auto- mation solutions with a template that others can fol- low to solve similar problems and waste. To read this entire article, which appeared in the April 2022 issue of SMT007 Magazine, click here. Feature Article Excerpt by Happy Holden I-CONNECT007 Journey to a Smart Factory