Options

Data format in predictive maintenance wizard

cleblanccleblanc Member Posts: 1 Contributor I
edited December 2019 in Help

This question is about the Predictive Maintenance wizard

 

I am running rapidminer studio 7.1

 

I would like to try the Predictive Maintenance wizard and I have acces to real time series data with sensor data as well as machine ID and faillure occurence.

I want to know what to put in the CSV file in order to load the reference data:

 

Looking at the rcontent of «//Samples/Templates/Predictive Maintenance/Reference Data»

 

The data is formatted with 27 column of 137 rows

 l2016-07-15 11_48_14-reference - Excel.png

 

I want to what to put in my input CSV:

Are M_0001 to M_0136 different machines or a time series

Are Sensor_1 to Sensor_25 different sensor or a time series 

 

 

 

 

 

Tagged:

Answers

  • Options
    Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    Hi Cleblanc,

     

    The Machine Id's are unique machines, they are not in a time series. The sensor data is also unique sensors, they are not in a time series.


    It's quite easy to do if you want to use time series data, you just have to install the Series Extension, available on the marketplace for free.

  • Options
    sunnyalsunnyal Member Posts: 44 Contributor II

    Hi,

     

    How do I run this predictive template for our data? where do we change the excel file? also does this need to be in the belwo format? what additonal data is needed? please guide me

     

    large.jpg

     

     I have 3 years of history containing all the possible sensor data reads of equipments, along with the corresponding weather details. I also collected failure history of the equipment. Ideally there should be some correlation between an equipment failure and some other factors like weather, power (ac/dc), fan speed,  etc.,   I'm struggling to figure out how to proceed with this predictions.

     

    Any advice on this is greatly appreciated.
     
     

Sign In or Register to comment.