How to prepare Multiple Sequences with two input vektors for Neuronal Networks?

Hi, I have the following neuronal Network with an closed Loop:

The purpose of the Network is to predict the Remaining useful lifetime of Rolling element Bearings. So i collected time series data from several bearings. For example:

Bearing 1:

 Features_B1 = [Temp1, Temp2, Temp3...; Crest1, Crest2, Crest3 .....]

 FFT_B1 = [Freq1_1, Freq1_2, Freq1_2 ...; Freq2_1, Freq2_2, Freq2_2 ....]

Bearing 2:

Features_B2 = ...
FFT_B2 = ...

How do i need to format the data to ensure that the Network is able to understandt that this are several sequenzeses.

As far as i my Research goes i need to use the following functions:

   catsamples % to Combine several squenzes together 
   preparet   % because of the open Loop???
   Input = {Features; FFT} % to Combine Input 1 and Input 2 into one Feature ???

Is that Correct? Which order do i need to use? How does this Change if i add an additional delay for one of the Input Vektors?

Thank you so much in advance.