COMPLEX ADAPTIVE SYSTEMS AND BIOROBOTICSModule BIOROBOTICS
Academic Year 2024/2025 - Teacher: Paolo Pietro ARENAExpected Learning Outcomes
The module aims to achieve the following objectives, in line with the Dublin descriptors:
1.Knowledge and understanding:
The course covers the main guidelines for understanding, designing and building bioinspired nonlinear circuits and systems with adaptive capabilities.
It includes a laboratory software / hardware experimental part. The course also includes guidelines relating to the design and implementation of
neuro-control models for biologically inspired robots.
2. Applying knowledge and understanding:
Students will be able to acquire and apply, in line with the degree course in Automation engineering,
skills in the analysis and design of nonlinear dynamic systems with particular reference to bio-inspired and artificial neural systems,
and to apply emerging skills in the control of complex dynamics aimed at handling; the relevant aspects also concern the ability to design and implement adaptive and learning systems for innovative perceptual machines inspired by the brains of some living beings taken as model organisms.
3. Making judgment: students will be able to choose the most suitable methodology for the realization of non-linear dynamic systems oriented to motion control.
4. Communication skills: Students will be able to communicate the results of their training both in Italian and in English, thanks to the specific training given.
5. Learning skills: students will refine the ability to learn and elaborate concepts in relation to the project of
complex dynamics oriented to biorobotics directly both from the lectures and from the material provided, all in English,
so as to be able to place themselves in a particularly advantageous way of in front of laboratory-type interviews, once the o
bjective of the degree has been reached.
Course Structure
Frontal and practical lessons, demonstrations with multimedial material; laboratory
Detailed Course Content
Introduction to Biorobotics and interdisciplinary aspects; in-depth study of non-linear dynamics in biological neural systems; biological neuron modeling and phase plane study; models of synapses and their modulation; computationally efficient models of artificial and biological neural networks; simulation examples with reference to case studies. Biological neural paradigms for the generation and control of locomotion systems; the Central Pattern Generator (CPG): in-depth study and comparisons in relation to animals taken as reference; implementation of control locomotion paradigms through non-linear systems and circuits (analog and digital implementation); study of real examples of biologically inspired robots, controlled by models of biological neural networks (implementation of "worm-like" wave dynamics, implementation of CPG networks and decentralized controllers on hexapods and multilegged robots). The role of complex dynamic systems in modeling and in the control of perceptive dynamics, with application to biorobotics: study of complex dynamics for the control of perceptive systems for biorobotics. Supervised and unsopervised artificial neural networks: applications to modeling nonlinear dynamical systems
Textbook Information
1. “Neuronal Control of Locomotion: From Mollusc to Man“, G. N. Orlovsky, T. G. Deliagina and S. Grillner;
2. “Dynamical Systems, Wave-Based Computation and Neuro-Inspired Robots”, P. Arena ed.
Course Planning
| Subjects | Text References | |
|---|---|---|
| 1 | Introduction to Biorobotics and to its interdisciplinary aspects; detailed study on nonlinear dynamics in biological neural systems | dispense |
| 2 | biological neuron model and phase space analysis, models of synapses and of their modulation; | Dispense del docente |
| 3 | computational models for biological neural networks; simulation examples referring to cases of study; | Dispense del docente |
| 4 | biological neural paradigms for the generation and control of locomotion patterns: the Central Pattern Generator (CPG) and the decentralised control | LIbro 1 |
| 5 | implementation of the locomotion control paradigms through nonlinear circuits and systems (analog and digital implementation), | libro 2 |
| 6 | examples of bio inspired robots controlled by models of biological neural networks: implementation of undulatory worm-like locomotion patterns, | libro 2; dispense |
| 7 | implementation of CPG networks and decentralised controllers on hexapod, quadruped and biped robots. | libro 2; dispense |
| 8 | The role of complex dynamics in modelling and control of perceptual systems for biorobotic applications. Toward an insect brain computational model. | dispense |