Testing part 1:

Monday I brought the prototype to the gym and had a lot of issues! When players ran across the line it produced a false positive every time. I  believe this to be because of the code. The upper limit to differentiate between a ball and a person is 100 ms and the lower limit is 1 ms. When a foot is running over the line it shifts the line a little bit forward. This is because the line is a prototype and  isn’t pasted to the ground like it would in a normal game. I believe that shift back in the line is affecting the FSR’s to give a short high output. I hope that by changing the lower limit up to 10 ms the issue will be solved. The fastest duration of impact for a volleyball I have found is 15 ms. Therefore, I believe the new lower limit will improve the false positive without missing any ball impact.  Another issue I had while testing was that the line stopped working all together while I was testing a hard hit. I believe this was because there is a recovery time for the force sensitive resistors. If an FSR continuously receives high force after high force, the FSR will not go  back to its high resistance as fast as needed. Therefore there is a recovery time for the two foot section before it can work properly again. I counted this recovery time to be between 30-45 seconds. Ultimately this is not a big issue. In the game of volleyball it is extremely rare that a two foot section would continuously have hard impact over hard impact. So the 30-45 seconds is a small enough window that this issue won’t largely affect the accuracy of the system.

Testing part 2:

After changing the Arduino code and allowing enough time in between tests, the second testing round on Wednesday of this week went much better. Testing four different scenarios: hard hit, free ball/bounce, run and 2 foot jump, I only had a percent error of 10%. The errors only came from false positives from a run. I believe this is a combination of the line shifting (from testing part 1) along with the fact that the wires are outside of the line. If wires are tugged when a player runs, the shift of the wires may create a false positive. All in all, I believe my false positives are due to the fact that the prototype is not as well constructed as it could be. Therefore, a professionally made line would not encounter such issues. Meaning the in and out line monitoring system would be even more accurate then it is now. Below are before and after photos from testing as well as a link to my a video of my prototype in action.

Free ball/bounce BEFORE:

Free ball/bounce AFTER:

Hard hit BEFORE:

Hard hit AFTER:

 

Attached here is a movie of the line monitoring system in action!