Project proposal: Limb-troller

Early on last semester, I originally planned to work game design into my project for this semester. When I met with Prof. Galanter about this, he encouraged the possibility while reminding me that the project would still have to put emphasis on the physical/interaction side, and not lose that work to the game design side. I shelved that idea since I wasn’t sure at the time how I could work physical computing into my gaming design bent.

Then this semester I looked closer at the phenomena of wearable computing. And I remembered some posts I’d read about how some people at Valve Software, a private video game development company, are experimenting with wearable computing in relation to gaming. Most of the posts I’ve seen come from the link below, in paraphrases or quotes. The whole article itself is rather long, but the second half deals more specifically with wearable computing, or rather the author’s decision journey towards investigating wearable computing. Last post dealt with the “what” and some of the “how”. This link deals with the “why bother?”

Why Wearables?

I propose to create a wearable gaming controller that will send control input based on the motion and orientation of the arm (for me, the left one as I am right handed). The game used to demonstrate this will be as simple as possible; I’m picturing a sphere flying/bouncing/rolling through space and obstacles. The simplest possible lighting and colors would be used, so as not to detract from the physical computing aspect. Basically a technical demo. I would like to get this “Limbtroller” (controller & arm portmanteau) to communicate wirelessly with the game if possible. Bluetooth type thing maybe; this is a less important aspect for the project than simply getting the controller to recognize the arm’s motion and orientation as useable input. Also the computer running the game should be able to receive the input and break it down into useable actions (for instance, an elbow angle could work in conjunction with finger movements to move the player sphere somehow, or perform some action). I’ve mentioned wireless Bluetooth as a possibility, since many computers can now accomodate wireless devices. I realize some infrared controllers can perform similar functions like this on gaming consoles, but laptop and desktop computers are my focus for this project.

So, design a game controller that can be comfortably worn on and operated by the entire left arm that will (wirelessly?) send signals to a simple demo game resulting in actions within the game. I would likely need trimpots, or some other sensor that can be rigged to read angle changes and angular velocity of arm joints. (I say other than trimpots because I noticed last time that the trimpots we had seemed a bit fragile for quick, repeated movements. Maybe something like a pulley; bending the arm changes the length of a wire/string that is read to determine an angle & angular velocity. This deserves further investigation.) The fingers would also be used, potentially also the connections between thumb and fingers. This could interact with arm position and orientation to create many different possible combinations of game input. Future games using this system or one like it could be developed. My first choice for the demo game itself would be to use the free version of Unity3D, which has many systems already in place to facilitate game creation, especially one as simple as this plan entails. The backup would be using Processing’s 3D capabilities, which while limited, do allow for basic movement. Aspects like collision detection and physics however, which would make the demonstration more engaging and entertaining, would be better implemented in Unity, which in fact already implements them with allowance for modifications.

A little note on why I’ve decided to use my left arm and hand, even though I am right handed: during games I control the mouse to aim with my right hand, but my left really gets most of the work out through the keyboard. This seems to be the default with most games. The left hand (if right handed) is expected to perform many actions and has many more buttons to choose from than the right. There are times when the amount of input required from my left hand exceeds what I feel is reasonable. This plan on focusing on my non-dominant hand my prove not as fruitful as I hope. In that case I can modify the controller to be worn and operated on my right arm. Parameters would need to be reversed or slightly modified, but this seems doable.

Due to the different aspects that must come together for this to work (Limbtroller design and building, game/character basic control, wireless? interaction, I can easily see this as qualifying for a two part project that can fill the entire semester. I’d rather have one big thing to show that works very well, than two smaller things that I wish I could have more time on.