Need idea for college entrance project

[AtomicWaste]

Sorry it's been a while since I've posted on these forums. I'd like to become a more active member of this community but I haven't had much time lately. Maybe now I will that schools out.

I'm about to enter my senior year in high school and that means it's time to apply to college. I'd really like to go to Worcester Polytechnic Institute - http://www.wpi.edu - and major in robotics engineering. It would be a dream come true for me as I'd be the first in my family to go to college and I'd be doing something I've wanted to do since kindergarten.

Anyway, WPI is test-optional. That means that instead of taking the SAT's I can submit alternative materials via WPI's Flexpath program.

"If you don't feel spending four hours on a Sunday morning is a good representation of your academic ability, we won't hold you to that." -WPI representative

. Don't get me wrong, I think my SAT score is high enough to get me in, I just want to be sure. So... That's why I've decided to design, build, and program ...something... and continually log my progress. Then submit the log filled with pictures, schematics, designs, ect. to WPI. I figure the best way to make an impression would be to show them I already know what I'm doing.

So what I need is a "something." Something the college admissions office would be impressed with. I am willing to spend anywhere between $300-$1000 on whatever it is and have the next three months to work on it.  I would like it to serve some purpose or real-world application. It doesn't have to be the next revolution in particle robotic propulsion, but it should be able to do something useful. It doesn't even have to be a robot per say, just something in electronics that would look impressive coming from a high school student. Please, I've scoured the internet for what seems like months trying to think of something. ANY advice would be amazingly useful.

Thanks so much in advance. I've learned so much from this place and can't imagine where I'd be without it. Stumbling onto this site was a dream come true. I've been fighting for opportunities my whole life and this place has given me so much. I love this place, I will probably never leave it, and can think of no better community to go to for help (something I absolutely hate asking for).

[Razor Concepts]

Read lots of research papers and see if anything catches your eye, and do some detailed study and analysis on one specific thing. Avoid just making a random robot that sorts candy or something. Also, for some reasons many adults think high schoolers reading research papers is pretty amazing.

[GWER57]

Try to focus on one specific technology. e.g. A robot that implements beacon navigation, or can recognize faces, etc.

You should probably try to have whatever it is do just one thing but do it well. versus a robot with a lot of technologies that doesn't do much.

Choose a specific topic and do heavy research on it.

And with your budget you have a lot of options  .

[AtomicWaste]

Thanks for the responses. Beacon navigation sounds like a pretty good idea and I might look into that a bit more. I was also thinking about RF technology. I thought of a simple low range, low data rate, ad hoc system but with a twist. Instead of each radio being able to transmit data only to devices in its immediate area, nearby devices could relay the data to other devices. For example. Say device A is 50 meters from device B and device C in 50 meters from device B and 100 meters from device A. If each device can only transmit data reliably at a fifty meter range, device A could not communicate with device C. But if device A could relay its signal through device B, then device C could listen to device A. It's a little messy, and There wold probably be a lot of interference problems, but it might work.
I figure I would create a simple PCB with the required transceivers, antennas, and MCU, all in one convenient package. Then build a few other simple bots that that this device could attach to (Via UART, I2C, SPI, recomendations?), and the device could create an instant communication network to perform some simple task.
I also think it would be pretty easy to find research papers on this topic. Everything from checksums to mechanics has probably been observed by someone.
Thoughts?

[Asellith]

Maybe some designated protocol for transmission like bluetooth or wifi will help with the interference problems or each bot will need to be able to change frequencies and other stuff.

Or crazy idea create two TX/RX channels and simulate the I2C protocol over the air by using the first channel as data and the second as the clock. That might get messy with delays and stuff. Sounds like a cool project. The biggest suggestion I have is even in failure a great report with possible solutions to the issues experienced and a lot of good test data would impress them a lot I bet.

one more note you might want to seriously look at a zigbee solution because thats what zigbee does in a nutshell is creating a mesh network like your talking. Maybe if time and cash allow take it to the next step after creating the functioning network and find a cool application for it. What ever you do just document it well that will leave the best impression on the admissions people.

[AtomicWaste]

Thanks, you've given me a lot of options to think about. I'm not sure which to try first. They all sound like great ideas. I especially like the idea of simulating the I2C. It might be a very easy way to control who's aloud to talk at what time.

I think I'm going to first try using simple RF modules like the ones that sparkfun sells. - http://www.sparkfun.com/commerce/product_info.php?products_id=8945 - http://www.sparkfun.com/commerce/product_info.php?products_id=8947. This will give a chance to truly build up a protocol from scratch. If I get them working,  I might move on to bluetooth, zigbee, ect. to further improve upon the protocol I've created.

When I'm finished, I'll be sure to post all my documentation, designs, code, ect. here for others to see and improve upon. SoR will definitely be leading my list of acknowledgments.

[Hawaii00000]

Sounds like a cool idea! Have you thought of making an automated paintball turret? I've never done this before, but my advice would be to document it  well an make it professional (look at Admin's wiring tutorial, use shrink tubing, and plan. )


Wire connector tutorial:    http://www.google.com/m/url?client=safari&ei=OgYZTKjyG-mvtwffmoyLAg&hl=en&oe=UTF-8&q=http%3A%2F%2Fwww.societyofrobots.com%2Felectronics_wire_connector.shtml&resnum=2&sa=X&ved=0CBAQFjAB&usg=AFQjCNF724r94lFO---mqKSZEciaU2sQNw

Here's a nice turret if your interested: