It is a very cool alloy.
A client of Acigan wanted piston like feet that self adjusted on a hexapod leg assembly a few years back. Nitinol pistons were used.
Basically what was described above is almost the same concept that was used. but attached at both ends of the piston/spring in your project. One end to base plate and one end to the fluid box. It may even last longer then the older motor type design since there would be three attachment points instead of only the center pivot used in the commercial type designs. they will either fail at the cantilever bar that does the rocking, or at the pivot (which is the weak point in the entire design) or the motors burn out due to the stress of the resistance the motion of the fluids produce that travels back down thru the cantilever, thru the motor shaft and into the motor itself.
So by changing away from that design completely and to a tri-point type design, the pivot weak point doesnt even exist, since all three attachment points carry the stress of the fluid motions and well, they are springs, they are coiled, so physics alone will reduce that stress travel and the coils absorb it just like axel springs do on a car or truck.
Hope this change of concept shows that anything can be done, you just have to climb out of the mainstream way of approaching problems and think of solutions that to some, appear to be out of the normal lines of thought.
If you go the Nitinol route… repost the results and let everyone know the outcome.
Good luck and most of all, have fun in the process.
.