What do you need to have an Industrial Internet of Things?
- Data
- Collection
- Analysis
- Feedback
Data. It all starts with the data. It has to be read from networked industrial devices like I/O, drives, and vision systems. It has to be integrated from non-Ethernet networks and devices. It has to be accurate, timely, and complete (see the post “Feed Me… PROFINET Data.”)
Collection. The data has to be collected and stored. But if it is just collected and not analyzed and acted upon it might as well be collected into WOM (Write-Only Memory). Yes, this is an old, old April Fool’s joke:
Getting and collecting the data is old news. PROFINET has been doing this for many years. And it provides a solid foundation for the balance of the IIoT.
Analysis. This is new. In the past we may have looked at graphed trends or alarm history or SPC. But the IIoT analytics needs are far beyond that. There is now a huge employment market for mathematicians and statisticians. They will be using more complex methods and looking to codify them into software like GE does.
Feedback. This is the part that makes it all worthwhile! Turning what has been learned into process improvements, uptime enhancements, and waste reduction. This calls for vertical industry expertise in addition to the mathematics.
You may be surprised that “Cloud” did not make the checklist. For me, “cloud” just means that a particular function is off-site. It doesn’t have to be off-site, but increasingly these days it is. Economies of scale, professional backups, and other benefits accrue.
It is understandable that many users are not ready to take all the steps outlined. But at least provide the foundation for future adoption of the whole IIoT by using PROFINET now. The data will then be available when you are ready to check off the next items.
–Carl Henning
[Cross posted at GE Intelligent Platforms blog: http://www.geautomation.com/blog/industrial-internet-things-checklist]