On Feedback Error Learning for Adaptive Soft Robot Control
Event details
Date | 03.10.2024 |
Hour | 11:00 › 12:00 |
Speaker | Niccolò Enrico Veronese, Politecnico di Milano |
Location | |
Category | Conferences - Seminars |
Event Language | English |
Abstract
This talk aims to explore how Feedback Error Learning (FEL) framework can be used to generate online learning
capabilities for soft robot control.
Soft robots are appealing in a wide variety of tasks thanks to their inherent advantages in safety, compliance, and
adaptability. However, accurate modelling and control of soft robots are still significantly challenging. For these
reasons we proposed the FEL architecture which is composed of (i) a data-driven model generating a feedforward
signal, and (ii) a feedback controller. The latter has two roles. Firstly, it corrects the action of the feedforward
controller when the tracking error increases. Secondly, it generates a learning signal to train the data-driven
model, allowing for online adaptation of the feedforward signal with respect to changes in the dynamic of the
system.
Biography
Niccolò Enrico Veronese is a fellow researcher on Robotics and Control at Politecnico di Milano.
He received the B.Sc. degree and M.Sc. degree, both with honours, from Politecnico di Milano, Italy, in 2021 and
2023, respectively. He developed his master thesis at the University of Oxford, UK, within the Oxford Robotics
Institute in close collaboration with the Control Group of University of Cambridge, UK. He presented two papers
on Control and Reinforcement Learning for soft robots at the IEEE Robosoft Conference 2024 in San Diego, CA,
and authored a review paper on Additive Manufacturing, published in Progress in Material Science. Niccolò is an
alumnus of the Alta Scuola Politecnica, an interdisciplinary honours program for 150 top students from
Politecnico di Milano and Politecnico di Torino.
His research interests focus on control and machine learning architectures, particularly their applications in
robotics, automation and soft robotics.
This talk aims to explore how Feedback Error Learning (FEL) framework can be used to generate online learning
capabilities for soft robot control.
Soft robots are appealing in a wide variety of tasks thanks to their inherent advantages in safety, compliance, and
adaptability. However, accurate modelling and control of soft robots are still significantly challenging. For these
reasons we proposed the FEL architecture which is composed of (i) a data-driven model generating a feedforward
signal, and (ii) a feedback controller. The latter has two roles. Firstly, it corrects the action of the feedforward
controller when the tracking error increases. Secondly, it generates a learning signal to train the data-driven
model, allowing for online adaptation of the feedforward signal with respect to changes in the dynamic of the
system.
Biography
Niccolò Enrico Veronese is a fellow researcher on Robotics and Control at Politecnico di Milano.
He received the B.Sc. degree and M.Sc. degree, both with honours, from Politecnico di Milano, Italy, in 2021 and
2023, respectively. He developed his master thesis at the University of Oxford, UK, within the Oxford Robotics
Institute in close collaboration with the Control Group of University of Cambridge, UK. He presented two papers
on Control and Reinforcement Learning for soft robots at the IEEE Robosoft Conference 2024 in San Diego, CA,
and authored a review paper on Additive Manufacturing, published in Progress in Material Science. Niccolò is an
alumnus of the Alta Scuola Politecnica, an interdisciplinary honours program for 150 top students from
Politecnico di Milano and Politecnico di Torino.
His research interests focus on control and machine learning architectures, particularly their applications in
robotics, automation and soft robotics.
Practical information
- General public
- Free
Organizer
- Professor Giancarlo Ferrari Trecate