Without power the platform rests in its lowest position (in my case sitting on the studs I added). When powered up, it slowly moves to the "neutral" position, which is half way between top and bottom. What bothers me when power is turned off is that the platform typically doesn't sink down symmetrically, but rather leans towards the side where the gear/motor offers the least resistance, as can be seen in the video. Not only does this look quite unprofessional, it can also lead to the struts touching and maybe scraping the motors. Plus, before entering the seat next time you probably have to manually move the rest of the motors all the way down, because typically they don't all sink down by themselves. This is more like a "crash landing" than a clean shutdown.

Here's how it works with my modified firmware. When power is turned on, the platform stays in its idle position until it receives positioning commands. It then softly moves to the neutral position, and then gradually adapts to the actual motion requests. When the simulation ends (no more positioning commands for 2 seconds), the platform softly moves back all the way down to its idle position. That way, when power is turned off, it is in a stable position that doesn't change uncontrolled any more, and is ready for the next start. Now that's what I call a smooth landing! This demo was produced using FlyPT Mover to control the platform.

I took the original SMC3.ino sketch for the Arduino UNO R3 and modified it to behave as described. See my Firmware page on how you can install this open source version of the firmware.

With these changes implemented in the firmware, all there is to do is to close the simulation software and wait a few seconds before turning off power to the platform. In my case I use a master/slave multiple socket outlet, where the PC is the master, and the platform (together with everything else) is the slave. When I turn on the PC, the platform gets turned on automatically, and when I shut down the PC, it gets turned off.