After taking spectra on the beamsplitter, and looking at a comparison between damping on and off, I realized the damping was not providing the intended effect.
I double checked a few things before starting to blame other crews :) :
-damping filters loaded and activated
-watchdogs untripped
-model correctly wired
-signal is going to the dac

When looking at the current and voltage channels at the ouput of the coildrivers when sending a signal to the osems, it seems like the coildriver is not receiving anything.

Will discuss tomorrow with Richard to see if that's a simple cabling issue.