Katur's TUI Blog

Last Minute Problems...

2006-05-06T08:46:00.000-07:00

So we finally started integrating our pieces for the final demo and disaster stuck. Our integration testing revealed a huge hole in our design that we did not consider. We found that the magnetic fields that our magnets were producing were providing interernece with the RFID reader/tags. So much so that we could not place the RFID reader inside our shelf as originally planned. The We had to do some major reworking to get the smartshelf to still be able to sence what foods entered and exited the fridge and also work with our existing code base. We finally decided we would have a "loading" area where foods where loaded onto the RFID reader from a safe distance away from our magnets and then placed in the shelf. The shelf and the RFID were distanced just enough so that we would not have any interfernece. Because of this change, we also had to make changes in the way our code issued commands to the interactive mode on the handboardy so that we got the right magnetism to each container.

The SmartShelf...

2006-04-30T20:56:00.000-07:00

My group and I are working on the SmartShelf, a prototype for helping dieters stick to their diets.

The system is going to use magnets to provide users hepatic feedback when removing certain items from the shelf/fridge and give them a reminder which foods are not on their diet.

We started be experimenting with electromagnets. We want to be able to control the magnets using the handyboard
so that we can programmatically control how much magnetism each container gets. Our first challenge was to wire this up in such a way that it would work correct. We quickly found that the handyboard did not provide enough juice to really power the magnets as necessary. We then turned to an external power source, but we had trouble wiring the setup in such a way that we could control the strength of the magnets. After some research we found our handmade magnets too weak to get the job done. We ended up buying magnets from an external supplier. The magnets purchased were strong enough but we still were having lots of trouble wiring the system to give us variable resistance. We tired using resistors and potentiometers, but nothing seemed to give us the control and variable strength we needed.

One comment I would make for future classes is to not think so much out of the box on this project. I understand that thinking out of the box helps foster creative ideas, but thinking about the technologies we had to implement this project a little more would have kept us from trying to implement ideas that were too far out of our range. I am disappointed that we were asked not to worry about the technologies when first looking at ideas for TUI's. It personally caused a lot of frustration when we were having such a hard time implementing the variable resistance.

After lots of frustration we were able to wire some relays to an external power source and the handyboard to generate the variable resistance we needed for the magnets. Once we were able to get this going we started to implement the RFID section of this system. We decided to use RFID to sense which items entered and left the shelf/fridge so that we can control the magnets on the fly based on which items were in each container. This section was fairly straight forward. RFID setup was the hardest part to get right. Making sure all the classpaths are set correctly is the biggest challenge, but once this was set, most of the RFID section was fairly easy.

Our final problem to battle was the interaction between the handyboard and RFID. We had no drivers that would help us communicate between the two systems. After researching online we were not able to find much help. We did however come up with a way to communicate between the two. We decided to use java robot class to issue keyboard and mouse clicks to the handyboards interactive mode. This allowed us to mimic someone typing on the keyboard the commands we would need when items entered and left the shelf/fridge.

We are now working on putting the finishing touches onto the system. More updates to come...

2006-04-19T19:46:00.000-07:00

I am very upset right now. I just wrote a 2 page blog entry but Blogger decided to go into maintaince mode and did not save my post.

I will update later

Handy Cricket

2006-02-26T17:01:00.000-08:00

My team and I used the Handy Cricket for the first project. Links on Handy Cricket can be found here:

http://handyboard.com/cricket/

Setting up the handy cricket was very easy. The device was already connected to a motor, photosensor and a light. To setup the cricket, we downloaded and installed the logo software on a PC. We connected the serial port of the PC to the interface and then tried to input a test program to see if we could get the demo application to run.

Our first attempt failed because we didnt realize we had to power on both the interface and the cricket (we assumed the interface would be powered by the PC). Once this was complete our next problem was downloading the demo code to the cricket. We were not sure if the interface needed to face a certain way to download the code, but after a few tries we were able to download the code.

After experimenting with the cricket we decided to implement the Motor, the light and the Handy Boards on board sound ability. The logo code proved to give us some difficulty. We had some trouble using some of the conditional statements, but realized you can nest conditions using brackets [ and ] We also found that extra spaces in the code would throw a wierd error when we tried to download the code.

We also found that inputting the sounds was also difficult. We found a website that had ring tone sounds and tried to translate them to the logo code but found out that some of the notes did not translate to logo so we just used the closest note and changed the duration of the sounds.

Another pain point was the pressure sensor. We were not sure the values that were assigned based on the amount of pressure assigned. We tired to use the output ability that was documented in the documentation, but we could not figure it out after many tires and we could not find any other documentation on it. After some testing we found that 255 was no pressure and 0 was very hard pressure.

Overall the Handy Cricket was pretty easy to use and there was enough documentation online to do some fairly interesting stuff. Below is a picture of our final project:

Welcome

2006-01-29T11:56:00.000-08:00

This is my first post for Comp 150 TUI.