I’ve been on an analytics kick lately with PROFIblog posts Statistics or Engineering which concludes:
I think there is a huge opportunity for OEMs and SIs with deep vertical knowledge. They can leverage IIoT, PROFINET, and remote monitoring (and maybe even some magic statistics package) to add value for their customers and income for themselves.
And Analytics for Everyone which concludes: “But I say that the future belongs not to the sexy data scientist, but to the engineer who can apply his vertical knowledge in the new platforms like Predix and MindSphere.”
“What do you mean by analytics?” is an ARC blog post. [I stole their title.] It opens with
Analytics is a hot topic. It’s the key to becoming data driven. It powers the IIoT and I4.0. It is the secret sauce behind predictive maintenance and predictive operations software. It is moving into all manner of applications and user interfaces. It is a marketing tool, a supply chain optimization tool, and a design tool. But almost without fail, every discussion about using analytics gets bogged down because of confusion about what is meant by analytics.
The article then goes on to define many of the terms encountered when considering “analytics.”
Also from ARC, GE Aims to Embed Predix on Every Intel Device. The key phrase is “GE’s objective is to embed Predix on every single Intel device with the customer just having to activate it.” Predix is the GE analytics home.
From ControlGlobal, Analytics opens new possibilities. My main take away:
Collected data must be contextualized: collected along with information that allows it to be related to the asset, problem or parameter under study. Then it becomes possible to use analytics to detect and predict issues and visualize the results.
This is where PROFINET plus OPC UA excel of course: providing data and putting it in context.
Also from ControlGlobal, Self-service analytics for non-data-scientists describes existing and idealized tools for mining Big Data.
You should really read the above mentioned articles to educate yourself on the role of analytics in the Industrial Internet of Things and Industrie 4.0.
Back to my earlier assertion of a difference in expertise needed for analytics, my conclusion on statistics or engineering: it should be AND. When such a dichotomy is presented I am reminded of Built to Last: Successful Habits of Visionary Companies by Jim Collins and Jerry I. Porras. They make the point “No ‘Tyranny of the or’ (embrace the ‘genius of the and’)”.
Statistics AND engineering knowledge (plus industry knowledge) are the analytics answer in the IIoT. Analytics needs data and PROFINET provides the data from the source (along with quality and diagnostics information).