Human Rocket

To fly like ironman basically sums up my inspiration :slight_smile:

I think the idea is great for motion controlled gaming as it fits the control scheme perfectly. Only thing missing for me is a VR rig, including a harnass bolted to the ceiling to have your feet free to move :evilgrin:

Current progess, a stylized concept art of my basic model


Work plan

  • Create basic model [done]
  • Create ragdoll
  • Create controlling armature
  • Move ragdoll to follow control pose, plan is to rotate the limbs
  • Add boosters + forces
  • Test setup in zero-G
  • Add wind friction
  • Build test level
  • If it all works up to this point I can start having fun with adding cool stuff (sounds, HUD, 3D-TV view, character modeling, special effects, other controller types, airbags, level design, localized wind, …)

Hopefully all works out as planned and I can keep some steady progress going. Apart from level creation I think the concept is simple enough to have progress witht he few hours a week I have.

I am excited to see how the project developes, good luck (:

Leuk idee. Succes ermee.

Thanks for the encouragement. So far I got the ragdoll setup, and some basic start for the HUD to already help me with debugging.
Imagine the camera up-side down to get a feel how the ragdoll bumped into the bars…


Just a short update of what I managed this weekend.
I got some basic input of my Razer Hydra into blender, not really worth a screenshot. Main issue is to get a sensible elbow direction (I get perfect hand positioning + orientation but translating that to some nice elbow rotation that does not jitter every way is quite a pain … ). I guess it would be cool to have a kinect or something to play with as well.

So for next time I will probably focus more on actuating the ragdoll to follow the controlling armature pose, and hope I come up with something good for the elbow issue eventually.

Nice work, do you have a video of any gameplay?

Did not get around to make a video yet, and I guess the real “fun” will start when I get thrusters working.

So far I have a controlling armature which controls the pose that needs to be followed by the ragdoll. For debug purposes you can see it on the top left of the screenshot. I am using torque to rotate the limbs, and apply a counter torque on the main body, so pulling you legs forward would also make the whole body bend forward as well. The top-right armature shows the actual character pose in world coordinates. I expect some version of it will end up in the final HUD as orientation helper.

Bottom right shows a color coded pose, my plan is to render it to a texture and count the colors to see how much wind each limb will catch :evilgrin: (just have to see if python is fast enough for this)

I also almost got the elbow pointing sorted out in a way I expect to be robust and sensible. So after that it will be a small step to add thrusters and make everything spin out of control (literally)


So does any-one have ideas how hard it will be to control a human with rockets strapped to hands and legs? :stuck_out_tongue:

yeah.

have the thursters all aim at a point above the actor when hovering,

‘throttle’ = mouse wheel up/down

‘Tip’ = WASD = adds to a offset vector from above the actor

QE = Rotate actor

my Floating vehicles in wrectified actually do this (to hover)

Great project Wraaah. I think that the rocket in the feet and hands is a dangerous idea, but it can only be done IronMan style. Otherwise the legs and arms wont have enough strength to control flight.
In any case, great project!
For flight control, I’d do it like in martial arts : Have specific flight poses and call it ‘Tonderu-Do’… In other ways, it would be a mix of swimming/skydiving and mountain climbing.
But it’s your game, make sure it’s not too technical, just believable and fun!
Keep it up!

Engineering wise it has everything against it, but it is :ba: and that is only thing that counts!
I do want to have some basic approximation to physics but nothing really accurate. Fun is indeed high on my wishlist.

Regarding controls, I want to avoid a simple arcade flight type of movement as it would kill the concept. I think it should feel reckless, dangerous and almost out of control. I had not considered extensive flight poses so much for control, but it indeed sounds as a cool alternative that could be do-able if you don’t have motion control gear. Anyway it all requires some decent play-testing before I get there.

This would most certainly be best done in C, and relatively easy do. The tricky part is finding a fast way to expose the image to C without copying it via Python.

I will first check if a small 32x32 image will already break performance before considering any optimizations or alternatives. And I am not familiar enough with C to get anything done without a lot of effort :(. I did start working with the python Ctypes module to interface with my motion controller directly through the DLL. I would expect ctypes to be able and interact with C code in general …

Anyways first get the thrust controls working in a sensible way to build up speed before worrying about the resulting wind.

After some experimenting I find my current system too complex to control (ragdoll + 4 thrusters). So I will make a redesign which will have the player as a single rigid body, and also only keep the hand thrusters. For model I was thinking of a big exo-skeleton around shoulders and arms, so might end up looking like a gorilla or so :smiley:

Basically the sum of all thrust has to pass through the player center of gravity else you will start to spin. And this is really hard to control with 4 thrust points.

Also having the player as 1 single object will make it easier to fudge some physics.

So back to simplicity, starting with a new concept for HUD design :stuck_out_tongue:


(Yes I know a lot of complex arrows, but I believe it will make sense in the end, and I expect to need all for debugging)

So lets hope this 2nd concept will fly