Arnaud Klipfel

Welcome! I am a Robotics Ph.D. student in the Institute for Robotics and Intelligent Machines and the School of Electrical and Computer Engineering at Georgia Institute of Technology. I am working under the supervision of Dr. Seth Hutchinson co-advised by Dr. Sehoon Ha. I received a Master of Science in the School of Electrical and Computer Engineering at The Georgia Institute of Technology and from ENSTA Bretagne in December 2022. My Master Thesis work investigated model-free Deep Reinforcement Learning (DRL) for quadrupedal locomotion to generate more natural and smoother gaits. I also earned a French Engineering Degree (Diplôme d'Ingénieur) from ENSTA Bretagne in mobile, fields robotics, as well as a Bachelor degree.

I have experience in robot control using non-linear, linear control theory, and Reinforcement Learning. My research has focused mainly on model-free Deep Reinforcement Learning (DRL) for control so far. I have designed and deployed policies for quadrupedal robots in order to learn smooth and diverse locomotion skills. My research interests lie in field robotics. I want to make robots interact with the physical world more. Having learned good representations of the world, and their environment I want to study how to make robots exploit their environment to solve their tasks more efficiently, and more robustly. Using physics simulators and DRL good high level representation can be learned, that an Optimization based controller could exploit for more robustness. Current robots primarily avoid complex real-world interactions, manipulate simple, symmetric, weightless objects.... I strive to develop and write code, algorithms that can be replicated on different robots, and in different conditions. Reproducibility, resilience, and robustness are key challenges in field robotics. I am interested in delving more into model-based DRL for control, Field Robotics, and learning world models for planning and more physical interatcions. Robots have to embrace interactions!