Community news, upcoming events and general discussions
Threads: 36 Posts: 88
Get technical support from the community
Threads: 1427 Posts: 7608
Threads: 18 Posts: 64
Tell us how to make XDK better!
Threads: 36 Posts: 119
Share and discuss community member projects
Threads: 81 Posts: 350
Many plant owners don´t yet know where to apply sensors like the XDK. They don´t know which data to collect or how to use it to improve productivity or to monitor conditions. A first tool that would really help to understand how their machines work and which parameters influence each other would be an easy to use correlation analysis of the data that just has been collected.
Of course there are many cases where forces and distances are related to each other. Thats a no-brainer. But what if you find that an increasing amount of vibrations and heat production would correlate (and thus lead to increased wear)? Or maybe energy consumption and humidity?
It might be an easy next step to provide such features on the software side. No extra hardware is needed. The End-User could then apply some sensors to their machines, collect data and re-think about causes of failure for maintenance services.
Matlab for example, already provides some visualization options: http://de.mathworks.com/products/statistics/features.html
I accept the XDK competition terms and conditions
Hello Jost, thank you for submitting your XDK use case! You accepted the BCDS terms and conditions correctly and your idea Sensor based correlation analysis will participate in the Ideation Jam. Kind regards, Manuel
thank you very much for your submission! Very interesting topic.
Could you detail on a concrete example/use case where this was actually done? It would be very interesting to see such an example.
I don´t know yet an example where this has lead to significant knowledge about machine behaviour. But I have heard of a few industrial manufacturers experimenting with similar features right now. The initial position is always the same: Nobody knows where to apply (costly) sensors and what they could learn from the data regarding wear or OEE (Overall Equipment Efficiency). Although a correlation feature would be easy to implement from the software side.
A quite funny website, that I found a few years ago gives you a really good example how "reverse" correlation finding look like: http://www.tylervigen.com/spurious-correlations There you may find that the number of prople who drowned by falling into a pool correlates with the number of films Nicolas Cage appeared in. This might not be surprising for some people :D
this kind of makes sense.. :D
seriously, this is really interesting. I think this could really be beneficial for a lot of people after they collected a lot of data and don't know what to do with it.
If you stumble upton more information just let us know :)