Tomas’s Status Report for 4/10/22

I mostly spent this week ironing out a couple Jetson-related issues that we encountered.

• First, following the code refactoring from last week, I noticed a significant drop in performance (about at 20 fps drop). I narrowed the issue down some extra CPU overhead related to our updated graphical draw-over. I was able to get around this issue by turning up the power setting on the Jetson to its highest setting, and activating all available cores.
• Second, as we got closer this week to establish two-way communications between the Jetson and the Due, we discovered an issue with accessing the “/dev/ttyTHS0” port from our docker container. I spent a few hours experimenting with the best way of making this device available, and was able to get it working. This means that next week we will be able to test this communication channel.

I had also planned for this week to begin testing the accuracy of our object detection system rigorously according to our design report plan. Unfortunately, due to the time spent addressing the issues above, the university holiday, and my own midterms, I fell behind in this area. Luckily we decided to allocate two weeks to this process, and I plan to catch up with this area in the next week, now that some of these integration issues have been resolved.