Wii360... perfect example for RC?

[Tim Arnold]

Stumbled accross an incredible project this morning that interfaces a Wiimote & nunchuck into an Xbox 360. Basically it goes Wiimote -> Bluetooth into Laptop -> Outputs to project board through custom driver -> Project board with custom firmware -> PS2 gameport -> Adapter for 360 USB. He explains this insane project much better than I ever could, so here is the link: http://obsiv.spaces.live.com/Blog/cn...F394!177.entry

Essentially, a project like this could realistically allow you to control your bot with a Wiimote. Now, it might be impractical (have to carry around laptop; long development process), but its definitely possible.

[Lil' Lavery]

Apparently giving constructive criticism is no longer allowed on Chief Delphi...

[jgannon]

Controlling a robot with a Wiimote? It's been done.

http://www.youtube.com/watch?v=mj4lsNw3IG0
http://www.youtube.com/watch?v=0qEotHQgUsg

[AdamHeard]

Originally Posted by Lil' Lavery

What advantages would controlling your robot with a wiimote grant? It seems like an impractical solution that doesn't really offer any advantages over simpler solutions.
The driving of your robot (I assume) would be driven by the joystick on the nun-chuck, which would be virtually identical to driving your robot with a single joystick on an XBox/360, PS2/3, or PC controller.
I assume you'd use the motion sensing to control the robot's manipulator, probably most adaptable to an arm-type system. While it offers a different style of control and a more compact system than systems like this one, it doesn't seem like it would offer any greater levels of control or major advantages in speed or accuracy of your arm over similar control styles. While true, it is often hard to find a robot-centric control style for manipulators (rather than operator-centric), this solution is more complex than other possibilities (such as the earlier example, a touchscreen, or model arm).

I think you're entirely missing the point, don't lecture him on the impracticalities (he already admitted them).... It's just cool.

[Kelly]

Originally Posted by Lil' Lavery

Apparently giving constructive criticism is no longer allowed on Chief Delphi...

Why did remove your comment? It seemed perfectly reasonable to me.

[Lil' Lavery]

Originally Posted by Kelly

Why did remove your comment? It seemed perfectly reasonable to me.

I received 4 negative reps from it, apparently most of CD doesn't feel the same way as you.

[Kelly]

Originally Posted by Lil' Lavery

I received 4 negative reps from it, apparently most of CD doesn't feel the same way as you.

So? They're just dots. It would be one thing if your comment had been truly inappropriate or unnecessarily rude, but you're just folding to peer pressure.

[Bongle]

Just to take the neg rep burden from Lil Lavery...

It's a cool project simply for its amazing ability to hack together many disparate elements, but I entirely agree with Lil-Lavery: Using a motion-sensing apparatus such as a wiimote to control your robot seems a bit counter-intuitive. I understand that there are teams that did it this year (one at GTR even made homemade wiimotes), and that is certainly innovative, but from my experiences playing my wii, I can't see a real fidelity advantage over other methods.

Since the accelerometers are the 2nd derivative of what you presumably want to measure (the position), then you've got to do lots of computationally intensive and potentially error-inducing integration to determine your position. Building a little model of your arm and tricking it out with potentiometers would allow for direct manipulation of your on-robot arm from your OI.

Sure you could use the accelerometers to infer the exact orientation of the remote (as many games do), but that can be tricked into innacuracy simply by the remote shaking a little (lateral or longitudinal acceleration would be misinterpreted as the remote rotating).

It would be an excellant off-season project to learn about custom circuits, programming, and about electronics in general, but I don't see any great advantage over more conventional methods.

[Smaug]

yeah it was pretty cool some one from combots controlled there combot with a wii remote it got half way across the arena and lost signal so it wouldn't be a good idea to use in first unless u modified the remote and boost its signal

[pyroslev]

Originally Posted by jgannon

Controlling a robot with a Wiimote? It's been done.

http://www.youtube.com/watch?v=mj4lsNw3IG0
http://www.youtube.com/watch?v=0qEotHQgUsg

using it on a industrial robot and on the ones in first are two different kettles of fish. Provided it can be done logically with less gear and headache while conforming to the rules, the more articulated arms and turrets might get a boost and be competitive with the scale arm model I saw this past year.

The nunchuck is probably the more likely to have broad uses for turrets and launchers due to simplicity.

[JesseK]

I don't believe arm control with a Wiimote would be stable at all. I played Wii Sports at a party a few weeks ago, and I was always crooked or off or shaky. Combine it with jumping up and down when you succeed at something and BAM you just hit the ref in the head. It would definitely take alot of practice and discipline to pull off.

For driver control, well, I played Descent 1/2/3 and could fly off of 3 axes at once using a single joystick -- it was sick and nothing but fun. All of those wierd hellical patterns you get from a spherical coordinate system? Yea I could fly those. I guess I'm a little biased there, but I will be stubborn and say that a single joystick is the best way to drive a robot.

On the other hand, with a hard-mounted WiiMote on your robot, you could get some serious sensor feedback control if the RC/OI would ever support it. I've already posted that idea here for future games.

[whytheheckme]

Isn't there something about a positioned IR transmitter thats involved too? Or are we talking about using the accelerometers only?

Jacob