Meet Matheus F. Reis, PhD student at the MarineUAS program that is currently working on an algorithm that allows to steer a robotic vehicle along a specified path defined with respect to a moving target and, just last week, has put this on-going work to test.
Hello! My name is Matheus F. Reis, and I am a PhD student from FEUP and researcher at the MarineUAS program. I have a master's degree in Electrical Engineering with emphasis in Robotics, Automation and Control by Federal University of Rio de Janeiro. My research interests are Robotics, Nonlinear Control, Robust/Adaptive Control and Machine Learning.
Q: What are you working on LSTS?
A: I am currently working with professor A. Pedro Aguiar on Moving Path Following (MPF) control algorithms for robotic vehicles. The main idea is to steer the vehicle along a specified path defined with respect to a moving target, which can be another vehicle, a beacon or even a fish or shoal. The LSTS provides the ideal infrastructure to test the applicability of our ideas in real environments.
Last week operations
Q: In what way is it useful?
A: MPF algorithms can be applied for all kinds of vehicles, such as AUVs, UAVs and ground vehicles. They can be used for marine life tracking, convoy protection and surveillance, to cite a few. However, the current MPF algorithms do not consider the presence of disturbances acting on the vehicle, such as maritime currents, waves and wind. Currently, we are looking into ways of applying the theory of robust control to improve the performance of the MPF algorithms in these scenarios.
Q: What are the main challenges and how have the tests been going?
A: In my opinion, the main challenge was related to implementation itself, and being able to predict the limitations of my own code. I used DUNE, which is the onboard software running on the vehicles, used for communications, control, navigation and maneuvering. My first contact with the software was challenging, but I was able to learn a lot with the help of the people from LSTS. With respect to the tests from last week, I had some initial problems related to my code, but finally I was able to perform the experiments and gather the data I needed to validate the performance of the robust MPF controller.
Matheus F. Reis
Still from last week tests: