FTC Teams 24484 and 24485
Programming
Programming - the brain and nerves of our robot!
PROGRAMMING SUBTEAM
Our programming subteam uses Java and Python to code the robot, which includes creating an intuitive and assistive control system for the drive team, as well as efficiently and safely utilizing the electronics. Being a programmer requires you to understand how the entire robot functions as a system, so getting involved with the other technical sub-teams (build, design, electronics) is encouraged. Although prior programming experience is useful, the willingness to learn is more important.
PROGRAMMING PROJECTS
TeleOp
For added manoeuvrability, we chose to use a mecanum drive, keeping in mind the drivers’ ease of motion since the bot is fairly simple, such as by avoiding control arrangements that might make the driver accidentally make a mistake. We also slowed down some of the bot movements during the TeleOp period to make it easier for the driver to be precise.
Tensorflow
In general, the Tensorflow machine learning tool manipulates frames by creating randomized filters to detect patterns. By the process of trial and error, the model trains by recognizing the team prop and verifies its accuracy by using a set number of manually labeled frames. After the model had been trained, we verified our metrics and trained it more for better results.
April Tags
April tags are essentially like QR codes but for our robot. It provides our robot with essential data such as X and Y coordinates along with information such as the something. We used snippets of code, the April tag library provided by FTC along with our knowledge of Java, to piece together a program that would read these April tags and provide us with valuable information. We had a few hiccups along the way, such as the April tags being incorrectly detected, or identified. But through a bit of perseverance, we were able to fix all the issues we faced.