Biomimicry Gas Robot
As a long term project I took a place on a small student team for Professor B. Lorena Villarreal's research project. The project is a chemical detecting robot, which uses Biomimicry to recreate a "Nose", which would be used to detect gases in an area, and locate the source of them. It uses the example of a dog's nose, which takes in air and exhausts it in different directions to track scents, as well as having chambers internally to mix air samples. The project took place over multiple Summer semesters, each trying to make some progress in an area to get the robot closer to a completed prototype. The methodology for the robot all branched off of Professor Villarreal's research. Our team of students used the research to make a working prototype.
I worked over two summer semesters on this project, both doing different tasks for that time. The first semester I was assigned to do the hardware design and fabrication. This had me learning and working with CAD and 3D simulation software. The base model was already made by Professor Villarreal. It was my job to take the model and put it into an airflow simulation to see how it performed. and make any changes to improve the design. For that we used Autodesk CFD, which at the time was integrated into the now-discontinued Autodesk Flow Design. The second part of my task was making changes to the model in CAD in order to get the flow correct in the simulations. Once the model was correct, we 3D printed the model to begin working on the prototype. In addition to the printing, I also created and milled out the power regulation and distribution boards. My teammates made boards for the other components, like the sensors and MCUs, I edited the CAD designs to fit their boards, and we put it all together in a prototype. We took the robot from drawings and ideas, and made it into a functioning device.
In the second semester my task was changed drastically. The goal now became using Robotic Operating System (ROS) to simulate the robot and its environment, so we could then port it over to the actual prototype and have it work. My task here was to make the environment the robot would be working in. DigiPen did not have the facilities for us to make a gas testing chamber, so we would have to simulate it. This was the entire team's first time looking and working with ROS, so we spent most of it learning the ins and outs of the program and compiling tutorials for what we learned. Using those tutorials, the team of students next year will be able to continue where we left off to finish the simulation. In the meantime, I got good experience using ROS to make 3D environments, as well as learning how their Node, Topic, msg, srv, client, and server systems all fit together. I can set up an entire basic ROS network to start simulating any robot. I'm continuing to learn to use accompanying software like Gazebo to make even more detailed simulations of the robots we want to build.