Feel Three - Affordable motion simulator + giveaway

[traveltrousers]

Hey guys!

Since no one seems to be coming up with new ideas for an affordable motion sim for HMDs it's up to me

I'm raffling one + shipping if you sign up for the kickstarter announcement on feelthree.com

The Feel Three is modular, so you can build it how you want it and add to it later, it's stylish, so you don't have to hide it in the garage, it's safe, since the moving parts are hidden, it's awesome because you can have a huge range of motion and it's going to be affordable enough so that everyone will have one. You can have a sim, or a sofa, and decide how big you need it.


https://www.youtube.com/watch?v=J04cP97T590

The main components are aluminium plates that fit together in a large sphere, this rests on a base with several castors with omni directional mecanum wheels providing 3 degrees of freedom. A gyro in the 'cockpit' provides rotational and directional information so the Rift gyro is always 'level'.
https://www.youtube.com/watch?v=qa23W-HrzRE

I'm currently looking into getting the panels made and should be launching a kickstarter once I have a decent metal factory on board, either in the UK or China.

Once this is successful I'll be adding heave (up/down) as an optional extra to the base.

Head on over to http://www.feelthree.com for more info, pics and videos, and make sure you spread the word if you like the idea.

Cheers!

[lowv]

The sleek design is definitely a good selling point compared to existing solutions! What price range are you aiming for? Also, what are the acceleration/decceleration characteristics like?

[NeoTokyo_Nori]

Wow, I think I like it

I have been keeping my eye on DIY sim motion rigs (for driving and flight )
for a couple of years, and I have never seen anything that uses this type of motion-rig design!

The 6-DOF rigs that some people are working on, are ridiculously complicated and
require many expensive motors and motor-controllers. Yet they are huge, quite ugly and have very
limited movement range. i.e the cost benefit ratio is so low.

Now, this seems like a simple and elegant solution for providing a potntially huge range of motion.


The only rig that could equal the range would be this,
and it would be much higher in cost (size, space required, money, power, etc. )
http://bsmotion.com/index.php/en/products/2fx-simulator

So kudos to you sir

[lowv]

Wow, I think I like it

Now, this seems like a simple and elegant solution for providing a potntially huge range of motion.


So kudos to you sir

Yeah, seems like you could now strap yourself to the seat and do full rolls in Elite:Dangerous . Wouldn't that be cool? I'm only wondering about the responsiveness of that construction...

[WiredEarp]

Looks good and doable - but the issue will be how fast it can react/respond (faster is better for immersion), and also I can see a problem in that there is no vertical travel. Vertical travel is very important in things like flight simulators etc to simulate bumps, turbulence, etc, but even for racing games, the ability for it to generate bumps is probably quite important.

[NeoTokyo_Nori]

I definitely see this rig as something most suited for space vehicles, or underwater vehicles,
where the whole experience should be smooth, but be capable of a large range of motion.

It could certainly be used for flight very well, and I dont think that bumps and such are a necessity,
unless you are really after a flight training simulator...

As OP mentions, the responsiveness would depend on the amount of torque the motors can handle, and apparently you can add motors for more torque, which is nice.

Sure it can be used for driving too, but its a matter of how realistic you want the driving to feel,
since the "look and feel" of this rig is quite unlike a real car. I have a racing rig now (non-motion),
and would want to add motion to it, just for car driving, due to the look and feel aspect.

For me, the most important motion element for a driving sim rig, is the G force you experience
under braking into a corner. Good braking control is critical for fast driving, and physical force is an immediate and clear feedback for how the car is handling under braking. Its something that would be much better than visual cue only.

The bumps and such are really not critical for driving, although would add "immersive ambience".
Its more for a theme park ride type element to me.

Looking forward to seeing this develop

[traveltrousers]

The sleek design is definitely a good selling point compared to existing solutions! What price range are you aiming for? Also, what are the acceleration/decceleration characteristics like?

The whole point of making this is to pitch it as an affordable solution since there simply aren't any available, all the motion sims are just too expensive, $5k+ unless you make your own, and even then people are spending $1k+ on home brew platforms. So we can't be sure what the price range will be until we start manufacturing but we're hoping for under $1k for the base unit, that's the point of mass appeal. Plus since it's modular you can have more or fewer panels and motors depending on your needs or weight.

The speed of rotation will depend on how you configure your platform, so if you're a space trucker hailing ore in Star Citizen you'll be happy with less performance, if you're a racing driver you'll want more. Again, it's up to you and you can add more motors for better performance if you like.

[traveltrousers]

Looks good and doable - but the issue will be how fast it can react/respond (faster is better for immersion), and also I can see a problem in that there is no vertical travel. Vertical travel is very important in things like flight simulators etc to simulate bumps, turbulence, etc, but even for racing games, the ability for it to generate bumps is probably quite important.

Vertical travel is why motion simulators are so damn expensive, you need powerful and expensive motors to lift the user... but I'd argue moving someone up six inches isn't such a deal breaker for a good sim experience.

Heave is a 'vision goal', so something that we'll be keen to add as an option in the future but not something that is essential to start with. Beside, if you watch some of the videos of motion platforms peoples heads jerking around on the bumpy parts it looks quite exhausting. Imagine being in one of those for several hours, your neck would kill you.

There should be plenty of space for tactile transducers so the bumps can still be felt, but they'll be a lot subtler.

[WiredEarp]

Oh, I agree that heave is expensive to implement (although some people have come up with clever counterforce systems to reduce some of the load) - i'm just pointing out that heave is generally considered pretty important by flight simulator motion platform builders, as far as I remember (not sure about racing guys, but I suspect the feeling of a car going up and down is quite immersive).
I do like the simplicity of the design, I can see it would greatly simplify some stuff. It reminds me a bit of the old R360 arcade game. I'm a big believer in starting and continually iterating on it myself these days, so I think its a good approach. A heave base could be added at a later stage.

One other minor issue with this approach is that the faster you spin the central capsule, the more force is going to be exerted onto the tyres. You may have to work through issues involving traction of the wheels against the capsule when reversing direction abruptly in faster versions.

For me, the most important motion element for a driving sim rig, is the G force you experience
under braking into a corner. Good braking control is critical for fast driving, and physical force is an immediate and clear feedback for how the car is handling under braking. Its something that would be much better than visual cue only.

You are not going to get that with this or most other systems I can think of. Thats an actual decelerative force, so you can only really fake it by doing things like diving, banking, or quickly turning the sim to throw you in that general direction - it can't provide an actual sustained force on your body.

I'd love to get into this stuff myself, I just have other stuff on. There are a few forums out there that have some really super talented guys who build homemade flight sims. If you haven't already studied it,
edit: somehow missed the bottom bit out - which basically just said theres some good resources on other sites devoted to homemade sims.

[zalo]

Perhaps vertical travel isn't so important:
https://www.youtube.com/watch?v=jFiU_Qq5iu4

You just need to keep the entire assembly from falling apart!

[WiredEarp]

Ha that is pretty cool!! Yep, perhaps it can all just be faked with sophisticated enough haptics.

[traveltrousers]

You just need to keep the entire assembly from falling apart!

That's pretty insane :p

Give me a speaker to keep the weight down instead....

I've been working on the sphere traction and hope I can get my 'omnichain' to work instead of wheels, massively increasing my traction on the sphere.

[cadcoke5]

Of course, the aesthetics of your design are probably the best I have ever seen for a VR rig.

Regarding the physical aspects, though I do have some concerns. So, I wanted to share the following ways to test some concepts that may save you a lot of money and frustration. In general, if something can be tested a significantly lower cost than something like the final version, it is often wise. This is especially true of something very new to your building and designing experience.

My gut reaction is that being able to do vertical flips is not all that desirable. It is one thing when you are doing a loop in a roller coaster, when centrifugal force helps to keep you in your chair. But, your device won't have that, and it will only be your chair restraints keeping you in your seat. It may not be something a user will tolerate for very long. Even transitioning from upright to upside down for a moment will be prone to discomfort. You start with weight on your buttocks, then to the side, and finally the weight is on your shoulder and lap. It will cause rubbing on the sides, plus your full weight is being thrown onto the other areas.

So, I suggest first testing your seat and restraints. You just make that part, and have a few strong friends to roll it around 30 times with you in it, to get a feel for what it will be like. If you don't like it, there is no reason to proceed and spend a lot more time and money to make a complete sphere.

My next concern is that I think traction will be a limitation of your current rendering. A test doesn't even need to look like a dome. But, make the total weight realistic. The goal of the initial test is to test your ability to turn it like it were a race-car simulator. So, rather than a spherical bottom, you might use a pipe bender to make a large hoop, the same diameter as your wheel base. You will only be able to do yaw movement, but that should be enough to know how responsive it will be in that plane, and will be the absolutely max you can achieve. I think the roll and pitch axes will be worse, because half of the wheels will be trying to push the ball up away from the roller. I think that will tend to reduce friction.

An alternative test rig would be to obtain a spherical tank. Perhaps you can contact a tank company and see if you can just get one end of a tank shaped like a perfect spheres. If you are lucky, you may even get a damaged one and just cut off part of it.

Again, I emphasize that you should make the total weight realistic and the distribution realistic. Your illustration shows the user on one end of sphere, so their weight will need to be balanced on the other side by the display, plus what ever counter-weight is necessary. Your illustration also has more weight at the bottom. So, you need a similar counter weight at the top. It is looking that even without the shell, you may need close to 3X the weight of the passenger.

If traction is a problem, then it may be solvable by putting some sort of retaining system that applies pressure to the top.

-Joe

[traveltrousers]

My next concern is that I think traction will be a limitation of your current rendering.

Would this help?

With regard to being inverted, I agree, and if you read my blog http://www.feelthree.com/2014/07/possib ... and-ideas/ you'll find that I generally don't think it's something you need, but it will be something that people want, at least until they have tried it and then they'll realise it was a waste of time. Locking the sphere will be an expensive problem and you'll need to be wireless. 90 degrees (45 each side) will probably be sufficient.

As for weight distribution you need to check your maths... 3x the weight of the passenger?? Not even close. We'll only need a smallish counterweight on the other side as long as the user keeps their center of gravity near the center of the sphere. And the further out you put the counter weight the lighter it can be. The center of the sphere is the fulcrum, the user/platform is the load and the effort is a smallish counterweight. Classic lever mechanics.

[cadcoke5]

Looking again at your first illustration, I think I may have mis-understood, thinking that you wanted a full 360 deg rotation in each axis. In your last post you say 90 Deg, which I think resolves the inversion issue. If you don't have a heavy lid, that reduces weight.

Also, the view your message showed was from the back, which makes it difficult to get a sense of where the center of gravity might be, and I may be misjudging it.

So, the above reduces my concerns.

But, now that you show the chain underside mechanism, I am not clear about what causes the half-sphere to rotate in yaw? Your first post showed the omni-directional wheels, which even if there were only 3 wheels, can provide 3 axis of motion. But, If I understand your omni-chain mechanism, although you have 8 motors, they only seem to work in 2 axes total. Is that correct?

Also, I just reviewed some of the engineering involved with calculating fiction (my schooling in this is from several years ago). Providing more surface area does NOT increase friction. The formula is simply Traction = Force * Coefficient of Friction. The surface area of contact is NOT involved at all in the equation.

What the chain would do, is reduce the point loading, so that your spherical bottom does not need to be as strong.

You haven't mentioned your educational background. You are obviously a master with 3D design. Do you have the background to calculate the friction? If not, then I am sure others here will be happy to participate in that way.


-Joe

[traveltrousers]

Providing more surface area does NOT increase friction.

Of course it does, why do racing cars use slick tires? More surface area, better grip. Why do people reduce tire pressure for off roading? Why does a tank never get stuck? Why do tractors use huge wheels? I'm not building an equation.

"they only seem to work in 2 axes total. Is that correct? "

Not quite, if you look by the outer castors there are 4 motors and omni wheels for yaw. This could be reduced to 1 for rotational yaw drift if the user didn't want to spend money on yaw. Then the platform becomes a '2.1' degree of freedom version. I will add an omnichain for yaw to the design too for people that want it.

"What the chain would do, is reduce the point loading"

Only slightly, the castors take the weight and the omnichains push up on springs to allow for engineering imperfections (the sphere won't be perfect, I'm not NASA). A set of castor wheels costs next to nothing and can hold huge loads, why over engineer the chains, sprockets, bearings and frame to hold this weight when a $30 set of standard parts can do it for.... $30 :p

[NeoTokyo_Nori]

Questions.

How much does the omini-chain add to the cost?
It looks like it is a significantly customized piece of hardware.

Are you sure the sphere is able to maintain its structural integrity?
The frame and panels could become expensive if they need to bear a significant load.

So is the whole thing still going to be "affordable" ?

[cadcoke5]

I said" Providing more surface area does NOT increase friction."
You said, "Of course it does, why do racing cars use slick tires? More surface area, better grip. Why do people reduce tire pressure for off roading? Why does a tank never get stuck? Why do tractors use huge wheels? I'm not building an equation."

While real-life always trumps equations, the engineering equations for friction have been around for a very long time, and proven themselves. Engineering equations are what allows a designer to create a design that does not have serious flaws. Using intuition, and the trial-and-error approach to machine design can be a better choice for easy/cheap builds, but when the cost goes up, so does the cost of a design failure.

Here is a link to good explanation of the race car traction issue. I know the idea of omitting surface area from traction equations is very counter-intuitive, but it is real. There are other reasons for making a race tire big. E.g. you must avoid melting the tire from overheating. Tanks want to spread their weight over a larger surface, so their wheels don't sink down into soft ground.
http://www.stevemunden.com/friction.html

-Joe

[traveltrousers]

http://www.stevemunden.com/friction.html
-Joe

An interesting link. I will be sure to use rubber with high adhesion and strong springs....

[traveltrousers]

How much does the omini-chain add to the cost?
It looks like it is a significantly customized piece of hardware.

The chain would be a regular bike chain, or even a motorbike chain, so a small part of the total cost and cheaply replaced.
If I can find a suitable off the shelf roller then they should be available in bulk quite cheaply, same with the washers and bolts. Omni wheels or mecanum wheels are not cheap anyway, and finding the ideal size not easy, but with the right parts you can make your own and to your own size.

So we'll only need one custom component, the clip to hold the roller to the chain link. Something like this, a folded length of metal with two holes. Right now it's just an idea and will need to be properly engineered but produced in tens of thousands would bring the price down to pennies.



I will always be aiming for affordable, and since it's modular and expandable this is where the biggest market is. The first motion simulator cheap enough for your average westerner will sell a lot of product.

"Are you sure the sphere is able to maintain its structural integrity?"

Yeah, sphere's are very strong, we'll have lots of connecting plates to hold them all together and they're bolted together on 5 or six sides. The top loses some strength if it's open but we can add a brace or reduce the allowable rotation. If 3mm aluminium isn't enough then you can keep adding thickness, the metal is cheap and light. There will probably be a market for 'heavy duty' versions in the states too, although this will increase the cost...

[cadcoke5]

I suspect that omni-wheels will be cheaper than a specially made omni-chain. But, in regards to you chain idea, I will mention that there is roller-chain common in industry that have tabs sticking out of the chain links to attach things. But, I am not aware of any that have the tab rotated the correct way to attach your perpendicular roller wheels directly. Though, if you are going to have custom round plates made for the sphere, you might as well get a quote on making parts of a chain. You can even purchase the rollers and pins separately, to assemble your own chain for your first round of tests.

Here is a page of some of the options for tabs on roller-chain.
http://www.mcmaster.com/#roller-chain/=tgyocs

You might want to use double-pitch chain, to facilitate your side-way rollers, and to reduce the number of them needed. They have links that are twice as long. But, that would also require a larger sprocket. You can get the double-pitch links with tabs also, but you will probably need to find a company that specializes in chain to find that.

-Joe

[traveltrousers]

Best way to see it? In VR of course! :p

[NeoTokyo_Nori]

Best way to see it? In VR of course! :p

Yeah, I was reading your blog about it. Its a good idea.
Certainly fits the target audience !

Still not sure about the cost aspect however : P
So I would like to see it be as modular as it can possibly be,
So it could build my own internal cockpit frame, for example..

[traveltrousers]

So it could build my own internal cockpit frame, for example..

yes of course, it will most likely come with a mount for your own chair and there should be plenty of places to attach other gear around you.

[traveltrousers]

progress continues....

[Polux008]

Your design and idea are realy fantastic, i don't think my wife and my childrens will be ok to let me put this on my 50M2 appartment but i'm really sure this will be amazin, hope the best for your project

[traveltrousers]

Your design and idea are realy fantastic, i don't think my wife and my childrens will be ok to let me put this on my 50M2 appartment but i'm really sure this will be amazin, hope the best for your project

Thanks, remember you can sneak it in as a padded chair for the lounge...

[traveltrousers]

My how things have progressed

Kickstarter soon!