Space  Robotic Force Moment Sensor
While the SPDM FMS performs the best of any space ones, even it has to be used with operational restrictions, to avoid drift skewing the readings.  
The design problems for the space sensors:
Temperature Lack of atmosphere drives temperatures to be either hot in sun or cold in shade. Robots are for moving things around, so they go into and out of shade repeatedly. 
Force.  It can be inferred from strain or deflection. Such measures are affected by thermal distorsion.
Large dynamic range. Long robot arms & capture 'free flyers' generates large rotational forces. 
High Stiffness. Makes strain or deflection more difficult to isolate from thermal distorsion.
Drift Space operations can be slow. Slowly changing measurements are plagued with drift
6 dof  Shear and moment loading cases can produce difficult to distinguish effects
Our concept moves the measurement out of the time domain and into the frequency domain, which pretty well eliminates drift and dynamic range issues.  To implement that, we have developed a sensor that is non-linear with respect to load.  We use an internal segmentially poled ceramic element to drive  modes (above 400 hz) of the sensor.   We had built a 3 degree of freedom prototype and with some ESA money managed to test it enough to see that though things didnt work exactly as expected, the results were good  no drift and steady wrt thermal.  Eventually, with enough analysis, we understand why the results were so good, and are ready to proceed with a 6 degree of freedom system, targeted at space robotic servicing market.  
We have almost completed a simulation of the 6 dof system, and will be releasing it to interested controls people soon.  The design will complicate controls systems slightly, but should meet operational requirements such as high stiffness for large payloads.  The changes in stiffness due to non-linearity are smooth  no sharp corners. 
In addition to the technical challenges, there are management problems which seem to plague fms development.  Every time a new engineer takes up the problem, a new patent is born.
web links:
www.gve.on.ca/FMS.html
www.gve.on.ca/cctomm.pdf   :  tech paper with test results

