In the first major part of my thesis, I looked at how to use the dynamics of a robot to enable it to lift heavy objects. The idea is that sometimes, robots are not strong enough to lift a heavy object through shear brute force. In such cases I showed that, if we have the robot move a little more intelligently, we can greatly increase its capacity. These results were validated on a simple robot prototype I designed.
In the second major part of my PhD, I am exploring methods of using robot kinematics, dynamics and trajectory optimization to throw objects. The first goal is to throw objects the furthest possible and the next part will be to hit a precise target.
During my PhD work, I also got the opportunity to teach the Rigid Body Dynamics course at the undergraduate level. With 160 students, this was a trial by fire exercise that I thoroughly enjoyed and that confirmed my passion for learning and teaching.
Finaly, I had an undergraduate intern work with me to explore a few different avenues that I would not have had time to dive into on my own. This project, over two summers and part time durring one year, helped hone my project management skills and structure my PhD project a little better.
Before starting my PhD, I worked as a researcher in France. The project involved working on the locomotion of a humanoid robot (not The Terminator). This robot (Tidom) with a total of 12 joints in both of its legs is about the size of a four year old child.
My work was to build the control algorithm from the ground up to control all 12 motors synchronously to allow other researchers to implement their experiments on the robot. To do this, I designed a C program on a PLC to communicate with and command all the microcontrollers in each joint using CANopen communication. CANopen is a communication standard that is built upon CAN, and since the PLC we used only had CAN libraries, I built my own implementation of CANopen to be able to speak with the joint controllers.
My master's project