Forum Discussion
Full-Body Force-Feedback Simulator
Hi everyone, I would like to introduce our project as we are getting ready to publish more information about it and thought the best place to share it before we go live would be here and hopefully get some feedback... :ugeek:
Our company was one of the first backers on Kickstarter pledging for 10 development kits, partly because we have a large number of developers, but mostly because we really, really, really wanted to see Oculus get funded! Anyway, I digress...
My name is Luke (hello) and I am the CEO of the brand new company 'RiftSurf VR' and for the last year we have been working on a 'top-secret' project to make fully Oculus ready simulators.
We have built our first prototype machine and have started testing it. I guess it is worth saying that at this stage of development we have over-engineered nearly every aspect of the machine as we will be running heavy testing on it as we hook up the software, so it is certainly a beast of a machine, we lovingly refer to it as the 'Big Green Machine' or 'Vomit Comet'
Take a look for yourself at our first unmanned test run here:
(Note: it is running in full test mode which checks all possible rotations and positions)
The machine has a very complex control system at it's heart, which we will continue to develop further and will form the 'brains' of our simulation machines in the future. Control of the hardware is through a .DLL which we are producing.
The whole point of creating this machine stems from our understanding that with VR the brain receives mixed information from the body and eyes and can result in the feeling of motion sickness. We decided to create a 'seat' that could simulate all the physical sensations matched with the experience within Oculus.
This fully immersive environment can simulate nearly all positions and a great deal of G-Force and Force Feedback.
So... why am I posting this now you may very well ask? :?
I would love to hear from you guys and discuss the project especially as we're looking at using Razer Hydra for our control system and I have some concerns about the magnetic field; we are using electronic motors and a massive amount of metal afterall! :roll:
I would love to hear any suggestions or ideas for the kinds of experiences you would like to see on our hardware?
Thank you all in advance for your kind comments and feedback, next week we should have video of machine with a rider and Oculus... fingers crossed :oops:
26 Replies
- It's been a while since we posted anything here as we're still hard at work trying to get everything connected and working smoothly. However we ran into a problem and after banging our heads against the wall for a week or so I thought it might be a good idea to ask here... probably was the first thing we should have done ;
So here is the problem:
Consider that we are rotating the pilot in the real world to match the position of the vehicle.
Consider that this pilot is wearing an Oculus which is feeding positional and rotational information back to the software.
This is creating some serious problems for example: You look upwards in the Rift (directly at the ceiling) and at the same time we rotate the vehicle to a fully 90 degree upwards position (also directly at the ceiling). The result is that the pilot is looking up which adds 90 degrees to his X rotation from his VR position in the vehicle, but because the pilot is rotated in the seat he should be actually looking directly forwards.
This becomes even more complicated as we want the general alignment of the VR view to be aligned to the Virtual vehicle which does not rotate on ALL 3 axis - 2 Axis for 360 movement and 2 Axis for Feedback movements (side to side and up and down).
We've already tried some different solutions such as creating a gyro that Oculus would follow, but this runs into issues when inverted 180 degrees...
Any ideas or help appreciated... :?: :) - Optical tracker connected to the top of the chair pointing at a marker on the head of the rider? Full 6dof tracking and just don't use the rift trackers.
"KydDynoMyte" wrote:
Optical tracker connected to the top of the chair pointing at a marker on the head of the rider? Full 6dof tracking and just don't use the rift trackers.
Yeah I guess that 'Crystal Cove' approach would solve the problem, I was hoping that we might be able to do it in engine with some clever code trickery :)
Thanks KydDynoMyte :)- Actually a TrackIR or Free-Track approach from years ago. Maybe you could still use the rift sensors as a safety mechanism to make sure the head isn't flopping about from the rider passing out. :) Most projects have all the elements worked out except for actually making the big robot arm. You seem to have started with the big robot arm instead. Well done. But I am still in search of an affordable motion sim for the home. I am hurting my head trying to think about how your adjustable roll axis pivot point would help/hurt/limit simulating forces accurately.
- Wow, only just seen this post. That looks like a well engineered piece of kit!
The first thing I thought of when seeing the stress testing video was the potential for broken arms if caught between the chair and the main arm of the rig. I know it's not the final version but I presume you're going to add some sort of cage to surround the chair to alleviate this? Or would the arms be strapped in?
As to your tracking problem, the solution already offered sounds like the best bet. As long as you can make the Rift get it's orientation relative to the fixed tracker on the chair then it should work (as far as my limited knowledge says!). But I thought the orientation data currently comes from the internal tracker and the external tracker (in the Crystal Cove videos) was just used for position?
Looks amazing anyway! - Thanks Boff :)
I am heading over to the workshop today and will try and grab some updated photos. The arm issue you mention certainly made us worried too, so we have now fixed it, along with a number of other 'safety' concerns. Working out from Ukraine it is surprising how many 'Health and Safety' issues like this don't seem to come up ;)
As for solving our problem in the short term, we're constructing a virtual rig of the machine which handles co-ordinates and rotation, I am hoping that something will come out of this that allows us to offset the orientation of the headset and keep everything aligned...
Hopefully upload some new photos later :) - As promised a picture of our guard screen to stop arms getting hacked off. It's still a prototype machine and obviously not for commercial distribution, however we don't want our engineers or sacrificial test pilots getting maimed unnecessarily
- Please forgive the mess of our workshop ;)
Today we did our first set of pre-destined testing with Oculus, and wow! even at this stage the level of immersion and sense of presence is greatly enhanced. Thankfully at this point we are not inverting the pilots into fully vertical or even upside down positions, so the issue with positional tracking is almost undetectable. One thing we are already noticing is that the common problem of 'sickness or nausea' was almost completely gone... so it bodes well for the future.
I took some video too, hopefully get it uploaded in a few days.
If anyone has suggestions or ideas about solving the positional issue, please jump in on the thread :) - That's good. Getting your arm hacked off is a quick way to ruin your day.
- Especially when it's your wife in the pilot seat ;)
- As promised Test Flight:
..Please forgive our workshop mess :oops:
