In 2014, I made a decision to walk away from manufacturing after more than 32 years of immersion in the space. My manufacturing experience started as a controls engineer way back 1982, when the Allen-Brady PL2 and GE Series Six were the hot
technologies. Over the years I rode the technology wave starting with PC based HMI/SCADA, then came the process data historians, then came the higher level manufacturing execution systems that sat between the process data historian and ERP, i.e. Purdue model Level 3.
I walked away out of frustration about how slowly the space was adopting new technology paradigms like SaaS and advanced analytics. These paradigms were being discussed and explored but not adopted in a meaningful way back then. So, I left for the realm of outboard supply chain where these paradigms are the norm and had been for a number of years.
It was a great experience but after two years, manufacturing was calling my name, so I came back in mid-2016. What a difference two years can make! What I am experiencing now is how global and regional initiatives like IoT/IIoT, Industry 4.0, and smart manufacturing are driving awareness and adoption at the C-suite level of technology applied to the operational side of the business with data availability being the fundamental driver. I truly believe we are experiencing a manufacturing Level 2 and Level 3 renaissance that I have been anticipating the better part of my manufacturing career. I just never thought it would take this long.
It is definitely an exciting time in the space. The most jaw-dropping observation I have seen since coming back is how advanced analytics techniques like machine learning are making what was only dreamed about two years ago, practical. Machine learning is a form of artificial intelligence and is currently the most relied upon technique in the data science field. It is used by Google for its self-driving car initiative and voice recognition applications. I understand its genesis can be traced to the credit card fraud detection applications.
Also surprising to me is how the machine learning algorithms are not closely guarded intellectual property of a few global conglomerate companies. On the contrary, these algorithms are readily available on the public domain! They are finding their way into manufacturing mainstream. Case in point is the way it is revolutionizing the world of predictive maintenance.
What once required a highly skilled data scientists and asset domain experts months to develop, can now be done in days without data scientist or a heavy draw on the asset specialist. And the data findings are far richer and more accurate; new technology can self-learn and autonomously monitor for data pattern anomalies.
About the Presenter
Tim Goecke is director of enterprise application integration at MAVERICK Technologies. Tim has more than 30 years of manufacturing experience with expertise in operations technologies and automation controls. For the last 10 years, he has been heavily focused on the data collection and predictive analytics technologies and applying them to manufacturing assets via the leading technology OEMs. Tim’s expertise extends across continuous, hybrid, and discrete manufacturing.