Capturing 30 seconds of data at a time and then analyzing peaks and valleys is probably a safe bet. Now for some code critique:
You don't need the Merge Error on the left side of the loop. You end up splitting the error wire again anyway. Merge errors at the very end of the code, after the loop.
Straighten wires wherever you can. This makes it easier for us to read.
Most people prefer to see terminals instead of icons for controls and indicators. Icons take up too much room. Right click on an Icon and uncheck the option "View as Icon". If you click on the Tool menu (top menu bar), then click on Options, then Block Diagram, look for the option "Place front panel terminals as icons" and uncheck that sucker. I don't know why NI has to set this as default.
On the loop border where the error wires come in and where they exit, replace the terminals (square box) with shift regiters. Right click on the box and select the option to replace with shift registers. It will replace one and the mouse pointer will turn into a shift register with an arrow. Just move it to the other side and click on the border near where the corresponding terminal is. Do the same for the DAQmx task wire (maroon colored). The reason for doing this to the error wire is this. What if there were an error in one loop iteration. On the next iteration, that error would be lost because at the beginning of the iteration, the value coming in on the error wire has no error. The shift register forces the end value of one iteration to be carried over to the beginning value of the next iteration. The reason for the DAQmx task wire is to form a habit of doing this. If this was a For Loop, and it iterated zero times (it can happen), the value at the other end will be the default value for that wire. In the task wire, the default is NULL (not a task). You will get an error trying to close a NULL task. The shfit register prevents the error by carrying over the input to the output so it will not be NULL. Do this for all error wires, references, tasks, sessions, and such with all loops.
At the bottom right of the loop, you have a STOP function inside a sequence frame. NEVER stop a program like this, except in extremely rare cases. By stopping this way, the rest of the code, like the closing of tasks, do not get executed. Instead, just wire a boolean to stop the loop. Then the program can end normally. Perhaps you could put the code inside the first two frames after the loop. You can then eliminate the entire case structure. Wire the max reached wire through an OR function to the loop stop. When the loop stops, the frame code put after the loop will then execute. Then the program will end naturally.
I'm not sure what the code is in the small loop after the main loop. This code will run in parallel to the rest of the code. I hope you don't think that it will only run after the main loop. It will start running as soon as you press the white arrow run button because there is no data dependency (input wires from other functions). What will happen is that the intial values for min reached (and other locals here) will be used. The small loop will run once and then it will never run again. Meanwhile your main loop runs and the values for all of these locals are changing. The small loop will not run again, and those new values will never be used. Result is that the Save Position Min and Max are probably wrong. If you want this code to run only after the main loop has ended, you need a wire from somewhere inside the main loop to the border of the small loop. It doesn't matter where it comes from and it doesn't matter if it is used inside the small loop. What this does is that it causes the small loop to wait until there is data on that wire before it starts executing. Data won't appear on that wire until the main loop stops. Instead of using those local variables, find the wires for these values in the main loop, and wire those all the way out of the main loop and then into the small loop to where the locals are wired. You really don't even need a loop here.
This is enough for now.
