The past few days have been focused more on software than math. I am becoming more familiar with:
For my research I had to create a demo for the UUV_Simulator. To do this, I booted up and underwater world with a UUV in it. These launch files were mostly written for me so I didn’t have to do a ton there. The UUV has dozens of topics, so I experimented with them and used RQT to visualize active nodes and topics. In order to control it’s movement from the command line I published a geometry_Twist message to the cmd_vel topic with instructions. The cmd_vel topic published to the cmd_accel topic, which ultimately published to the thruster_allocation_input topic, which specified the required forces/torques required on the robot. It was pretty cool to see the thruster allocation matrix appropriatley power each thruster to acheive said forces/torques.
Everything that I was doing was visualized in Gazebo, so I got to play with the world and the models a little as well. Additionally, I used Rviz software to visualize the sensor data being recorded by the UUV. I liked seeing the raw image data (which was just a bunch of numbers when I echoed the topic) be turned into actual camera footage I could view.
Gazebo world running with Rviz displaying data gathered from left and right UUV cameras
For my Mobile Robotics Class we were given a Rosbag with some sensor data and asked to replay it, and figure out what the robot is doing. So far I have just opened the rosbag (contains one metadata yaml file and one database db3 file), played back the data, and gotten a list of all the topics