Cuddles's Blog

So, we went to schmoocon after a very very long night, and it was really really cool. The conference taught me a lot about what I want to do and how cool it is to be where I am. In high school people like us were always made fun of because we weren't “cool”, and to be in a conference where there's a thousand people that are interested in the same thing I am is pretty cool. Especially to see all of them pwning Earth.

All in all, I enjoyed this whole winter session experience and I learned more than I could learn in most classes. I can't wait for the next con.

Seacrest, out

Well this week has proven to be very fruitful for our project overall. My algorithm now is fully functional. It can count when one person moves straight through and when people come in through both sides of the sensor array. I combined my code with the two Robs and we now have a fully functioning program that reads if a person comes into an area and sends that to the router, also incorporated with Rob Rehrig's PWM code. The only problem we have is that the STOP mode doesn't fully work, so we need to figure out what's up with that.

On an extremely good note, we met with Motorola, Convergince, and a couple other companies I think but I really don't remember them. The meeting was a really cool experience because it was awesome to be around other engineers who develop really cool toys for us to use. In particular, one of the companies developed a program that uses cameras to provide a extremely formidable security system. The cool thing is our cameras would work perfectly with their software. I unfortunately was unable to stay for the entire meeting, but I heard it went extremely well. They even gave Burke the program that I was talking abut FOR FREE, which is awesome because now we have a GUI to work with.

Well I have come a long way since Friday when I couldn't even get the sensors to read 1 person. Now, with the introduction of flags and a wait function, the sensors read 1 person 90% of the time, and rarely read 2 people.

In addition, me and Rob3ar decided to do away with the TPM and KBI interrupts and instead place an RTI interrupt to activate the PWM every tenth of a second. My algorithm works with the RTI, and I can even get the stop function to work when the program reads that there are 0 people in the area. However, the algorithm still needs some tweaking. Sometimes it works and sometimes it doesn't, which is not good. 10% functionality will probably not roll well with motorola so I will optimize that.

Well today I continued to work on my algorithm, and I finally tested it with the actual infrared sensors. They were digital, and easy to work with my code. The main problem I was having was when the sensor should have been reading 1 person was passing, it was reading 3 or four. I still have to tinker with it to optimize it but I will be done hopefully by this weekend.

Well I figured out the stop function today, and now I have to code the counting algorithm to use it the right way. I will work on it a lot tomorrow and try to get something tested and done with.

The parts will be coming soon so I could do some real testing of what will actually be used. We had a meeting on Friday and we discovered that the two Robs and I have kind of claimed our own ports and interrupts. Rob Haislip has the receiving data interrupt and is using port B, Rob Rehrig is programming the PWM interrupt (which will have the transmitters blinking once every 10ms instead of on all the time so we can save power) and has Port C. I have claimed the KBI interrupt and am using Port A pins 4-7 because only those 4 can detect high edge levels.