Issue link: https://iconnect007.uberflip.com/i/911509
82 SMT Magazine • December 2017 expertise and tapping on the expertise of our customers and partners to be able to help com- panies get insights from the large amount of data they collect. Las Marias: Nowadays, equipment manufactur- ers are developing some sort of AI technology or machine learning into their systems. Why do you think that is so? Nair: Instead of just basing computer program- ming on a set of codes that the computer will perform—and it is limited to the instructions the computer is given—what happens is data is fed and models are created based on that data, so that decisions can then be made by the com- puter based on the wide varieties of data and scenario that are fed in. It's used in things like under- standing how autonomous vehicles would react in differ- ent conditions. Basically, by getting these large amounts of inputs and creating scenar- ios around them, the comput- er learns how to create deci- sions based on real-time in- formation. When it comes to the manufacturing side, it could be where people understand, based on inputs, of what kind of test requirements are need- ed to increase, for example, the yield. On the production side, it could be determining the yield conditions as products are being manufactured, and how one can im- prove the yield. These could all be use-cases for machine learning and AI. Also, robots used in the manufacturing line will increasingly learn how to work along with humans to be able to continually optimize. Because, really, the future is not just purely robotic lines, but how robots work along with humans in to increase efficien- cies—because there are some things that only humans can do. Las Marias: Many manufacturers have lega- cy systems in their factory lines. How can they transition to a smart manufacturing model Las Marias: Do you see manufacturers now adopting these technologies? If not yet, what remains to be the challenge? Nair: The places where I see quick adoption are in high-value systems, like aerospace systems. For example, people that make high-precision turbines—they are already a step ahead with regards to integrating their manufacturing to the IT level to do predictive maintenance. That seems to be carrying on as economies of scales take place, and people understand exactly how to use these smart machines, smart devices, the ability to connect to the cloud and to do ana- lytics, and to have artificial intelligence (AI) al- gorithms that can provide insights. As that hap- pens more and more, I can see the adoption go downstream also. Las Marias: Industry 4.0 in- volves automation, connectiv- ity, and data. Nair: Yes, and connectivity to sensors, not only connectivity among processing units. There is one more thing that I would like to add: it's also the abili- ty to do data reduction at the edge, close to where the sen- sors are. Las Marias: Where does Na- tional Instruments come in to help manufacturers in their Industry 4.0 jour- ney? Nair: National Instruments come in to the test side, providing the ability to do data manage- ment and integration of these data manage- ment to the enterprise level. We also come in very heavily in terms of predictive mainte- nance. We are also helping companies in the verification and test, because all the data col- lected can then be fed into artificial intelligence boxes, if you like, where there will be analytics that do these AI and give people a better under- standing of how to do predictive maintenance, better utilization, and so on. So, where Nation- al Instruments comes in is sharing our platform Chandran Nair INDUSTRY 4.0 AND THE PLATFORM-BASED APPROACH TO TESTING