I have been working on using the lower stage sensors (L1 and L2 OSEMs together with optical levers) to get information about the longitudinal displacement of each test mass of a quadruple suspension. I have been experimenting with trying to match up yaw signals on the quad because this degree of freedom is relatively uncoupled, unlike pitch or longitudinal. The results are somewhat promising, except for an unexplained factor of ~10 difference between the OSEMs and optical lever data. More information can be found in this report on the DCC:

https://dcc.ligo.org/LIGO-T1300957

The matlab code that was used to generate plots in that document is

/ligo/svncommon/SusSVN/sus/trunk/QUAD/Common/MatlabTools/QuadModel_Production/Diagnostic_sensors/generate_damped_yaw_models.m