I'm eager to try out some more with Microsoft Azure Machine Learning and would like to find a data set to make a use case concerning predictive manufacturing. Microsoft already offers a data set (semi conductor) for a use case like this, but I would like to try out some more. Does anybody of you know where I can find another data set similar to the one provided by MS?

Basically I'm looking for a bunch of sensor data in a manufacturing process and a classification whether it came to a failure or not. I have tried the UCI Machine Learning datasets already (it only features the semiconductor dataset that I have already used) and researched the Kaggle repositories as well.

  • I'd be curious to see how that turns out. Is there going to be a public Github repository or website?
    – tbm0115
    Commented Dec 9, 2015 at 15:11

2 Answers 2


Look into MTConnect. Basically, you can read from a stream of XML coming from any type of device that utilizes the MTConnect "standard". MTConnect is part of the manufacturing solution for Internet of Things.

In a nut shell, anything that has an electrical pulse can theoretically be monitored using sensors. These sensors can be tracked and the signals can be interpreted into meaningful data, which can then be transmitted using HTTP/TCP in the form of XML using the MTConnect "standard". As a developer, you can then access this stream using HTTP/TCP. Here is an example of the schema you would use and here is an example of a transformation of the output stream (using XSLT).

A lot of manufacturing machines (mills, lathes, etc.) can utilize this format, however the format is broad enough that you could setup MTConnect on anything from as simple as a coffee machine to something as complex as the International Space Station. It all just depends on the scope of your application.


See http://mtconnect.mazakcorp.com/ for live data from some machine tools in production at Mazak Corporation in Florence, KY. Details of each data stream are listed at that URL.

NIST data from their Smart Manufacturing Systems testbed is published in several forms. See https://www.nist.gov/laboratories/tools-instruments/smart-manufacturing-systems-sms-test-bed

For similar data streams, but simulated, see http://agent.mtconnect.org.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.