Results

Results of the Design
Discuss the problems you had implementing your design and how it worked. Problems like the CAD tools crashed, my workstation did not work, or I did not have my components are not design problems. Consider the following:

We successfully built a physical fitness machine that tests your ability to do sit ups and run the "shuttle run". Along the way, we ran into several design challenges.

Our original design used the SmartFusion, and had many extra sensors and peripheral devices, such as an mp3 player, pulse sensor, and disk memory. After re-evaluating our plan we decided to exchange the SmartFusion for a more powerful yet challenging processor, the M3. As a result, we had to make software modifications and spent alot of time configuring our radio and devices to talk to the M3. Since this was challenging and time consuming, we had to pare down on our sensors and choose the most important peripherals to make our fitness tests work. We eliminated the Mp3 and memory completely. Also, we had challenges with soldering devices and making filters. We fried our accelerometer, however we got a new one and made it work in our finished product. Also, our Wii Motion Plus sensor got shorted in the soldering process, so we were unable to use that in our sit ups fitness test(Originally, we used a combination of the accelerometer and the wii motion plus to determine when the user completed a sit up). We solved this issue by taking more data from the acceleromter and modifying the algorothm in the software.

Also, we had to change processing and control implementations to accommodate sensor limitations. For example, the light break sensor only worked for very short distances when we used the accompanying laser. As a result, we had to find a stronger laser, and alter the code to work with the new circuitry(account for different voltages, steeper drop-offs, "bouncing" from the laser, etc).

Although we fixed the majority of our hardware and software issues, there are still two unresolved issues in our design.

1. The encoder is still not perfect. It is pretty smooth but some of the time the bouncing causes it to turn left when you are actually turning it right. We added bouncing code but it was not intricate enough.
2. The pulse sensor detection was very difficult to work with. The output signal of a “pulse” was weak and the change in voltage of when there was or was not a heart beat was too minimal to determine when there was actually a pulse. The sensor consumed too much of our time.
3. The gyroscope was never implemented as a second check for our completed sit ups.

Fighting Childhood Obesity....For America!