Sunday, September 20, 2015

September Open House 2015

Another UConn Open House for prospective students has come and gone. Thankfully, this one didn't involve any show-stopping calamities. The biggest problem was that I had only gotten 3 hours of sleep the night before and, as a result, my focus was obviously impaired. This meant it was very difficult for me to reliably control the drone with my mind as I alter my focus for different commands like taking off and spinning.

The day started bright and early as I showed up to ITE a little before 9 AM to help move stuff over to the Jorgensen for our department's table. Interestingly enough, they had made a new poster for the CS/CSE/CompE departments and I was actually in a couple of the pictures on it.

After hanging around for what felt like almost an hour waiting for students and their families to start showing up, I went back to ITE to set up my drone demo. After putting on the headset and opening up my API for interfacing between the EEG and drone, I noticed for some bizarre reason it was only sending the gyroscope data so I couldn't do much of anything, including make it take off or land.

For once it wasn't the drone being temperamental

I was frustrated for several minutes but then remembered I still had my ad-hoc program to fly the drone with the headset although I wouldn't have as much control over it since I wouldn't be able to access the gyro data. It was good enough since we had a short time limit for the presentations but I still wasn't terribly happy I had to use the old program. It had worked fine back in May and I was convinced I hadn't changed anything about it since.

The department tried their best to be systematic and punctual with the tours and timing, yet we still encountered widely different time spans for presentations and sizes of tour groups. At one point we had a 20 minute gap in between groups, and the groups ranged anywhere from 3 people to over 30. People seemed to enjoy the presentation, though and I had a few curious people ask questions about it and my past internship at Google.

By the time I had finished with demos, everyone else was long done and ate almost all of the many boxes of pizza that were at Castleman. Thankfully, they still had a full cheese pizza waiting for me.

Although I'm a senior now, this was by no means my last Open House since there's another one coming up in October and one for accepted students in the spring. Hopefully by then my API will be functional once more... or maybe it was just a sleep-deprived brain causing all the trouble.

Tuesday, September 1, 2015

The Gameplan

Senior year at UConn has already kicked off and we're wasting no time getting ready for senior design. We have a solid team of 5 formed and a plan for one heck of a project: a mind-controlled RPG-- or a demo of one I should say.

We plan to use an Emotiv headset like the one I used for my mind-controlled UAV project to use the main character's psychic abilities. The plot isn't fleshed out yet, but the demo will cover up to 4 different mental actions (the fewer, the more reliable) and most likely take advantage of mood detection, built-in gryroscopes, and/or facial expression detection. We are even considering using a suit capable of picking up user movement allowing them to control the character with their body instead of buttons altogether to provide a more immersive experience. Using a Kinect is a possible alternative.

If we actually include motion tracking, this is going to be one heck of a demo

Nothing besides the group is set in stone yet though, as projects need to approved of by a committee since not everyone can do an independent project. There are company and university sponsored projects, but none of them seemed very appealing to us in comparison to our game idea. We and other professors we've already pitched the idea to believe this will be an innovative project introducing new paradigm of gaming instead of yet another typical game a group of seniors decide to throw together. 

As for the thesis, I will be figuring out the plan for the next component tomorrow. I already have some ideas after working at Google for a better implementation than the one I had in mind before this summer. Hopefully this third of the thesis goes smoother than the last one.