This week, I focused on making sure the integration between the Jeston and Arduino worked and finalizing the mechanical design so that we can officially start building this following week. As of last week, I was only able to get 1 value sent from the Jetson/received by the Arduino. The Arduino was working correctly in response to the specific value it received. However, when testing a string of inputs from the Jetson, the Arduino seemed to only output the result of the first input, making it seem like there was some buffer issue. Eventually, I was able to fix this, however, the transmission+receive time is a little too slow, and was something that I continued to work on for the rest of the week to improve. On the mechanical side, a few things occurred. The acrylic piece that we had ordered for our previous door design arrived, the Professor had provided us with wood pieces for our frame, and I talked to people at Techspark regarding purchasing plywood from them. The plywood pieces at Techspark ended up being too small and thin for our purpose, while the acrylic seemed to look like the better option, with some tweaking with the wooden pieces. With all of this in mind, I once again modified the mechanical design to now use the acrylic piece as the lid and door, with wood pieces on each side to create for the axle to then be attached onto using clamps. The overall acrylic lid will be connected to the wooden frames. I made a few drawings (shown in the team report) and decided on measurements that we can then provide to Techspark next week so that we can start cutting out our pieces and start building.
We are on schedule, prepared for the interim demos this following week, with the software and hardware subsystems somewhat operating successfully (though the integration of getting actual model output values rather than a script being sent to the Jetson still needs to be implemented). While there are a few issues with our implementation that we will need to go back and resolve (ex. Further fine tuning and getting proper outputs from the model in a format that can be sent to the Arduino), the individual parts are showing some form of operation, which we hope to improve upon in the next few weeks as we continue to integrate.
My next steps for the following week are to successfully complete the demo, start mechanical building, and help resolve the format issue regarding the model output and Jetson+Arduino integration.