Senior Project

USF Senior Design Project

and Hardware Design Competition Entry for the IEE Southeastern Conference 2002

This competition celebrated the anniversary of the original video game, Pong.

The goal was to use a video signal from a camera suspended over a table to automatically track and hit a white ball.

I teamed up with my good friend Scott to build such a system, which placed 7th out of 28 entries.

This was the highest ranking USF had ever gotten at one of these competitions, and I found out why.

While some major engineering schools had teams of 10 to 30 students and budgets ranging from $5,000 to $20,000,

USF gave me about $800, which bought the camera and built the table.

The rest of the hardware came from school, home, or was purchased with my money, about $1,200 out of my pocket.

I bought a special video card for my home PC to receive the camera video feed.

My USB web cam had software for automatically updating a photo bitmap file with the video feed at 30 frames/sec.

I wrote a program that continuously scanned the photo bitmap for the white ball.

It would build a track over time to predict where the ball would be when it reached the paddle.

The paddle would then drive for the predicted position.

This system could handle a ball bouncing off one side then heading for the other corner.

The paddle would actually move in the opposite direction knowing the ball would bounce back.

I had to use such an advanced solution because the system was not quite fast enough to simply follow the ball.

This system was the only system of the entire competition to predict the ball's future position :)

The PC would send the predicted position information to a programmable logic controller (PLC).

The PLC was programmed to use air valves to move the paddle to the ordered position.

The paddle had an electric eye that triggered a pinball solenoid to hit the ball as it approached the paddle.

The design had two limitations that held it back to 7th place:

1) A 5 HP air compressor was needed to keep the paddle moving.

It would trip circuit breaker of the power plug in the competition area.

So, we had to charge up the air in another room.

It would run out of air before the 5-minute matches were over and get too slow to intercept the ball.

We had inquired about the availability of power before the competition, but were told it would be fine.

2) The interception calculations did not account for the fish-eye effect of the camera.

As the ball would approach a corner, the distorted camera image would cause the paddle to move out of the way.

It just so happened that when the ball was served, it would tend to go right for one of these corners.

If by some miracle the ball stayed away from the corners and the air pressure was still high enough,

the system was almost unbeatable.

UF won the competition with an electric drive system that was so fast, that it did not need to be smart.

It simply followed the ball and never missed. Amazing what money can buy :P

If we had been able to identify and correct our two problems prior to the competition,

I believe our system would have placed 2nd or 3rd. Oh well, got an "A" on my project for graduation :)

Many professors from other schools noticed the unusual compressed air setup we were using.

No one else had anything like it, and the noisy air compressor always drew a crowd.

What had made us choose to use compressed air instead of a precision high-speed electric drive?

My personal credit card and the El-cheapo USF engineering department!