When Jason Prater, VP development at Plex said that the data the company collects through its cloud manufacturing suite can predict monthly US automotive employment you could have knocked me over with a feather. I've been arguing since forever that the data collected by cloud operators is far more useful and valuable than the application itself. I frequently ask why software vendors don't go all out to both gather and analyze selected, aggregate data. The answer I usually get revolves around some vague notion of confidentiality and/or privacy but I'm not buying that.
I'm aware for instance that one vendor was able to predict the 2008 downturn as a result of analyzing flows through their system. Another vendor has told me the SIC code system doesn't work for their customers and is actively developing its own version so that it can provide finely sliced data to all its customers. In the UK, the tax authorities have been collecting data against which to marshall investigative probes into small business margins. I have no doubt that US and other governments' agencies are engaged in a variety of tactics designed to better understand commercial behavior. That still begs the question - why are we not seeing more of the same from vendors?
Plex didn't have a broad range compelling answer although Jim Shepherd, the company's chief strategist offered the view that measurements of many kinds are familiar territory for manufacturers. That makes sense and explains why, when all the talk about so-called 'big data' as a new thing is getting hyped, Shepherd's response is terse and on point:
We've been doing big data for so long we don't call it big anymore. It's just data.
I like that response and especially so when we see it in roadmaps for the company's intelligence plans. The image above is of an early design stage concept of what Plex would like to deliver and which was floated for discussion during the analyst session. While this screen focuses on the financials, there are others in the pipeline for manufacturing and supply chain. Assuming Plex gets this acceptance tested then it is precisely the type of information I'd want to see as CFO. The only reason not to want this is the fear of failing and yet I often wonder how many businesses would be alive today if only they knew what was generally going on around them? This is a firm step in the right direction that I hope others will recognize and follow.
Here's the kicker. Users don't have to do anything to get this. It's just there. That's disruptive for the business intelligence community. How good is that? On to machines.
Last year, Plex surprised us with a hands on factory floor demo which had several of us 'making things' and which made extensive use of sensors and barcoding. It was a fun way to show us what Plex can do to make the factory floor simpler and more fail safe.
At last year's PowerPlex, CTO Jerry Foster ruminated on using Google Glass on the factory floor. This year they went a step further.
In another powerful demonstration, Plex showed us a prototype ruggedized Glass tool that improves safety and efficiency by avoiding the glancing up and down at clunky barcode readers. Swipe gestures transmit data back to the Plex systems in real time with almost zero operator latency. That's a significant advance. An alternative came in the form of a finger scanner and who could not be mesmerized by the bubble graphs of output on a global map that turn Plex data into the nearest thing I've seen to Candy Crush Saga.
All of which speaks to an exciting future for the increasingly automated plant floor. Of course there are still the cultural hurdles to overcome of looking a tad outlandish but as a few of us quipped: plant people are used to looking like dorks.