This past week, I placed a significant focus on the construction of the robot. We had most of the materials (extrusions & acrylic boards) by the beginning of the week; however, we had to order new connectors which took a bit longer to come in. Regardless, we finished constructing the frame of the robot and are looking to nail in all of the acrylic boards by the end of this week. I worked with Ella on the construction, and we spent around 6-8 hours to set up the frame and the materials for the innter mechanisms.
On the side, I finished setting up ROS on the raspberry pi. Although this was a task I had worked on in an earlier week, I had to run through the steps (flashing + installation of ROS) on a different PI. As mentioned in a previous report, the PI 5 is incompatible with ROS 2 (few work-arounds that I do not believe work well), hence I set up ROS on the Raspberry PI 4. In addition, I have been researching ROS packages (ex. Moveit, to use for motion control. By the end of this week, I would like to have completed the code for obstacle avoidance on the Raspberry pi. There are a few interesting ways to use ROS (ROS wrapper that works with ultrasonic sensors) to accomplish this, and so I will be working on integrating these components before the end of this week.
Jumping off of the other ROS packages described in the Team report, I also did research on Isaac ROS and April tags. The Isaac ROS VSLAM (Visual simultaneous localization & mapping) package looks very promising for our project – it takes in rgb-d image data to estimate the position of a robot relative to its environment & start position and produces results that can be used in navigation. If we can integrate VSLAM well into our project, we might not even use the Raspberry PI.
Our team is slightly behind schedule as construction of the robot has been taking longer than expected. However, we expect to finish the majority of the build by next week. I will also be working on the motor control code.
0 Comments