6DOF parts back from paint!

Hi I'm one of the people working on this platform (girl in the pictures).

Just downloaded Mover to check it out. Looks like it would work quite well for us. But seems that you haven't fully implemented the UDP output yet, is it something you are working on? That would be the easiest way for us to integrate quickly as we already have our firmware set up to receive UDP message from a custom test game.

Thanks for the great resource

I'm working in version 2.9.
Right now it broadcasts to all ip's, but only for one network board.
In 2.9 it will send also to all boards.

Do you need something specific?

@pmvcda We've got mover installed and just trying to sort out the UDP stuff, will post questions in your thread as the come up. Happy to share a video once we get it setup!

@Zed
-Ya, you can certainly get air in it now, I don't have the math on top accelerations yet but regardless yes, it can move you around pretty quick if you ask it too! We also haven't fully unlocked the motors yet (I think we have them at ~80% max at the moment) but when at full beans they can each draw 40amps so ~1 kW each or ~6 kW total.

That ~260W number was what the unit was burning total right then and there, IE rocking someone around. At rest power burn should be very close to zero. The big thing about our motors is although they can be position or speed controlled, they are at the base level force controlled motors (with position/speed/etc being stacked ontop of the force control). So essentially what happens is the motors detect a load of whatever and then just keep upping the air pressure until power draw is nill, then you are good to go. That's how we can move someone around (gently) and only be burning a few hundred Watts, as for performance curves etc, we are still collecting data for just that, but in theory you could have a stupidly heavy platform, still burn very little to move it (gently) but at the expense of lower max acceleration.

Bounce depends on how tight you set the control loops and if you want them to be stiff vs compliant is just a variable. That being said adding some predictive compensation from air compressability is on the to-do list. There is however some flex coming from the chair mounts (stiffening those up is also on the to-do list.)

As for the sales side: At the moment the platform is a scaled down (power/force) tech demo and still needs work to get it to a standard product. we do sell our motors mainly for force feedback applications and we have a few clients using them to build platforms (robotics applications). Depending on the interest we get on this thing once its launched in a more public forum we may/may not push for either building them OR just supplying the sub components. Really the main imputes on the project wasn't really to make 6DOFs but to show an example of what our motors can do in a somewhat more interesting way and attract some more attention.

And also thanks for the kind words!

When I tried to test it out by sending to a Unity game on the same machine that listens on the port I specify in the UDP output (IP 127.0.0.1) It appears to send only while I press the connect button while the % is going up to 100 then stops sending, then sends again while I disconnect while the percentage is ticking. So basically I'm not sure how to maintain the connection, does it need to recognize a network board in order to establish connection/continue sending?

For the platform, our controller is a single board that has an IP address and a specific port it is listening on, so if I understand correctly the current version might work with our set up. I will test it out on Monday.

Thanks @WalkerYYJ, good answers! I’ll be keeping an eye on this. The abrupt/sharp movements are where I saw the bounce but the smooth movements @Rebecca McWilliam was riding didn’t show the bounce and would be great for flight simulation. For the stuff I do I wouldn’t see much bounce but interested in what you guys come up with.

Thanks again!

It's a bug, the UDP was not tested in the 2.8 build as it should (most of us use serial over USB).
2.9 is corrected (I hope). I will try to send you a build tomorrow for testing (it's late here).

Amazing !! And I thought my new BLDC .motors were responsive and have detail, pfffffff

I know you might not sell them to the public but how about to a few keen DIY'ers ?

Good luck with the rest of the project

So to understand the tech behind it, it uses pneumatic cylinders to hold the load (air pressure) and the magnetic linear coils offset the position up or down.

Does this mean the coils temporary hold the load on position until the pneumatic pressure increases or decreases to hold the load in the new position?

In the video I can actually see these "steps" the coil motors trying to hold the load on their way down fighting perhaps against the building air pressure that varies by the load? Is the pneumatic air pressure distributed to the top/bottom chambers by servo valves, or just on/off single speed relays?

So basically these are pneumatic/magnetic coils actuators combination, not just magnetic, right?

What it looks like to me is that the magnetic coils are compensate for the low accuracy of pneumatics, filling up for the small accurate positions needed.

Anyway, looks impressive technology! Well done!

You got it! You could of course do the whole thing with just the magnetic side but it would drive the cost and power up significantly. So its a bit of a hybrid approach we are taking. We had originally considered to use magnetic or coil springs, but since we were already using pneumatic tube for the project anyway thought we would give compressed gas a shot first. There's still quite a lot of testing/tuning and redesign underway so we certainly haven't settled on whats going to be the best way to control the gas side, however variable flow rate valves are next on the list for testing.

@SilentChill The motors we built for this unit are probably still a bit undersized for what someone would want for a deployable system. Give it a few more months and we shall see where it ends up!

Hi all, it's been a while but thought I'd post a small update. We've deployed a few of these now and after ~3 years of tweaks (and a few new generations of motor designs) we are now commercially launching kits containing motors, controllers, reference designs, etc.

Fair warning, these are priced for high end OEM & Lab applications (~50K-25K depending on volume). The motors are all made in house with domestic supply chains so are unfortunately not cheap to build (yet). But if you have an application that needs tight, fast, low latency, & low current draw then drop us a line and happy to chat!


https://irisdynamics.com/products/raven

Great!
Is it still an hybrid gas and magnetic design?

Yup! We have (and include) a little control box that lives in the base and automatically adjusts pressure across 3 zones (front, middle, and rear motors) to keep the system balanced. You need maybe ~30psi (~2bar) for a heavy rider and total volume of consumed air is extremely low. For stand alone air sources we have been sending units out with these: https://airsquared.com/products/scroll-compressors/p09h017a-bldc-c/ or you could 100% just run it off of a small construction compressor / existing shop air.

We have a client using these for vehicle mounted sensor stabilization and they are just using a small bike pump.

We have a few more videos and articles planned over the next few weeks, will be getting into a bit of a deeper systems dive in one of those.