Some Good News

Well, relatively speaking. It turns out that Emotiv had some problems with their shipping company raising prices, which left some packages stranded and others in a backlog in their factory. They only recently discovered the degree of the problem, and have transferred to a different shipping company they've had better luck with in the past to clear out their backlog. Hopefully this means I'll have my headset shipped in the next few days and have it here to work with over the break. It would have been nice to use it in testing while I was here during the semester, but I'll take the small favors where they're granted.

The backbone of this project is pretty well put together at this point. It's looking like most of what I'm going to need to do is hooking up the headset to my software and then doing a lot of fine-tuning on various aspects including the flight stabilization I mentioned in a previous post and the detection thresholds on the headset's readings. As usually seems to happen, the stuff I thought was going to be complicated turned out to be fairly simple. The complicated stuff is a whole list of steps that I didn't know were even a part of the project when I set out to do it!

Regardless, this project should make for an entertaining semester, and I have to keep reminding myself that at the end of it, I'm going to have a helicopter I can fly with my mind! How cool is that?

And Failure

Coming off the high of last week, I'm experiencing a bit of a low right now as it pertains to this project. it has past the two week estimate for the outside length of time it should take a headset to ship and it seems mine still hasn't. I'm hoping that there isn't a problem with the product stock because of holiday sales, though I can't imagine this is an especially popular Christmas present.

On the more immediate home front, I also have an issue that has come up with the library I'm using to control the drone. Apparently, much of its flight stabilizing behavior is part of the control apps, rather than native to the drone's on-board firmware. This means that hover has a few kinks, as the drone drifts a bit more than it should, and it doesn't compensate for things like wind when being controlled via the computer. Apparently, I'm going to have to program this functionality into my control program myself. Oh well, it couldn't be smooth sailing the whole way through this project.

Successes

This project is actually coming along significantly better than I'd anticipated it would. "Mind-controlled helicopter" is not something that I'd thought would go this smoothly from the outset, but so far things are coming along really well.

Emotiv's headset apparently takes a while to ship, but in the meantime, I've discovered that they provide a free emulator, so I can use that to test my code. I was a bit concerned that transferring commands from the Emotiv software to the OSC converter, then to the Processing sketch and from there to the AR.Drone would wind up being an overly complicated process, but I actually wound up figuring it out in about fifteen minutes, twenty on the outside.

I can now successfully use the Emotiv software to tell my drone to take off and land. I'm excited for the headset to arrive so I can really get this thing working.

Processing...

I stumbled upon an unexpected solution to my problem of how to get the commands from an Emotiv EPOC EEG headset to my Parrot AR.Drone quadricopter this weekend: Processing! I've used Processing in several of my classes over the years and have some proficiency with using it, but it isn't something I'd consider a go to development tool, and, in fact, isn't something I've frequently come across outside of those classes. That said, I've found a library that allows Processing to connect with the AR.Drone and send it piloting commands.

It also seems that Emotiv provides free software to convert their headset commands to Open Sound Control packets, which Processing can receive and interpret using another library I've found. I finally have a good handle on how I can put this project into action, and I find it more than a little bit funny that I'll wind up using Processing at the core of this particular project.

My Quadricopter Has Arrived

I've gotten my AR.Drone, and it actually works even better than I'd anticipated.

I've had a chance to test it both indoors and outdoors at this point, and while it does have the necessary control level to navigated the somewhat cluttered interior of my apartment as seen above, it really has an amazing range of motion and control that stands out when given room to truly maneuver in an open field.

Now I just need to figure out how to connect it with an EEG headset to give me mental control over it.

Parrot AR.Drone Quadricopter

I believe I have found my helicopter!

Parrot's AR.Drone Quadricopter is typically flown using an iPhone or iPad app. It's on board sensors include a camera, accelerometer and sonar that allows it to hover and maintain very precise positioning relative to the ground. It allows for precision control in all directions: up, down, left, right, forward, backward, and rotation both clockwise and counter-clockwise. If I'm going to have a mind-controlled helicopter, i think this is my best bet as a base to work off of.

Thought-Control Helicopter

Following up from my previous post, I've been putting some serious thought into what my capstone project for this year will be, and I'm considering taking my inspiration from a video I saw a year ago. Specifically this one, which depicts a man using an EEG headset to get a basic toy helicopter to take off by thinking the word "Fly."

I think I'd like to do something similar, if a bit more advanced, and try to set up a thought control system to fully pilot a toy helicopter. It would need to be a slightly more advanced model than the one in the video, which looks like one of the cheap toys that doesn't give you much control over it, but I think this would be an especially interesting project. I had mentioned previously that I was interested in where thought recognition technology might be headed, and what better way to find out than to go there myself!

Brain Computer Interfaces

I have a significant interest in BCIs and their on-going development as a viable technology. In the past year I've learned that EEG-style control headsets have become appearing on the consumer market place, and I find this development pretty fascinating in terms of its implications for where technology is headed.

The two most prominent companies that I've seen are NeuroSky and Emotiv, and I think the products they're turning out are pretty exciting. Thought recognition technology may still be in that relatively basic, somewhat clumsy state that voice recognition was in a decade ago, but it's still much farther along than I might have anticipated, and the manner in which voice recognition has developed in recent years makes me very interested to see whether something similar will happen with BCIs.

First Post

My name is Chad Griffeth, and I am a senior interactive Multimedia major at The College of New Jersey, looking to work in the video game industry after I've graduated. My time is split between creating simple Flash games and creating 3D models and animations in Maya. I also enjoy working with Photoshop, both modifying images and creating graphics from scratch.