Some good news and some bad news.

Good news: Our technical paper was accepted.

Bad news: DARPA said that they decided that they're only going to let 20 teams enter the race so they'll be checking the teams' vehicles in early December to determine which ones are best. They never said anything about this before, but we'll be one of the twenty.

I made a sketch of what our rolling chassis will roughly look like.

Tim from VIA posted an interview of me at MobileRobotics.org.

I decided to use make a 3D laser scanner for the Expeditor.

It's the same as radar except that it uses lasers (ladar). It will be used to create a three dimentional map of the terrain in front of it. I'll be able to use one of the cameras that I already have to sense the reflection. This will give you a basic idea of how it works:

The laser will shine through a lense that will split it into millions of laser beams that are all on the same plane. My friend has one of those lenses. This would make it so that if you shone the laser on a white wall there would be a horizontal red line that went across it. The camera will be mounted above the laser, and will filter out all the light that isn't at the laser's wavelength. This would make it so that it could only see the reflection of the laser beam off objects. The software will be able to tell how far away the object is by its y-position on the camera image. This is only good for sensing objects at one elevation, however. The entire scanner will be mounted on a tilt mechanism that will let it scan up and down. If you still don't get it read this - http://www.seattlerobotics.org/encoder/200110/vision.htm. It's a very very nicely written article describing how to do this.

There are two things I could do. One is to replace the camera vision with the laser scanner, and the other is to use both. There are a couple difficulties with using both. First, I can only use one webcam at a time. I can connect them both to the computer, but it takes about three seconds to switch from one to the other. There are a couple problems with using just one. First, for the laser scanning it has to filter out all light other than the laser light. This could be done at the speed of light using a bandpass filter, but then I couldn't get a normal camera image unless I moved the lense out of the way of the camera. I could filter out all other light in the software, but that would be more time-consuming. The question is, if I have the laser scanning information, do I need a normal camera image?

Using a laser scanner (ladar) would not prevent me from using the neural network software that I've already written, but it would allow me to use that software along with new boolean-logic software to determine when an obstacle is in the way.

Sent our final technical paper to DARPA.

I also finished the accelerometer, or bump-o-meter. It has a battery connected to a potentiometer. The battery is only used as a weight, however. When the accelerometer is mounted to something that's bouncing, the battery bounces up and down, and the potentiometer can measure the bouncing. This is so when Expeditor is bouncing too much it will know to slow down.

Made a new sketch of final vehicle, this time with a swing arm.

Optima Batteries agreed to give us a 50% cut on one of their batteries. We still aren't sure if we want to use it - it's very large and heavy.

We got an email from DARPA responding to our technical paper. It was rejected, but we were told exactly why. The only reason was that we didn't give enough information in several categories. We will make sure to give all the information possible in our final tech paper.