We are trying to predict electric motor failures/maintenance,so started reading about ARIMA model for predictions.
In every example i noticed data series consists of independent observations at different points of time. Eg temperature series / sale numbers / stock prices etc etc, so the data series looks like as follows:
Time1 : Value1
Time2 : Value2
Time3 : Value3
.. and so on …

In our motor data set,we do not have independent values, meaning when motor starts, its current goes from 0 to say 5A max and then slowly comes back to 0, so in each motor run we have say 6 values.

Time1 : 0 : 1.5 : 5.1 : 3.3 : 0.8 : 0
Time2 : 0 : 1.3 : 4.9 : 3.2 : 0.9 : 0
Time3 : 0 : 1.2 : 5.2 : 3.1 : 0.8 : 0
… and so on …

so at each point of time i have a series of data points. I am really not sure how can i model this series of series of data using ARIMA.

I obviously have no prior knowledge of ML, AI and everything in between.

I will appreciate any help or reference to relevant resources.
Thanks a lot!

Just googled ARIMA and started understanding its datasets.
Need a pointer in correct direction or best model to analyse series of series data.