Exploiting Soft Materials, Biomechanical Structures, and Neural Mechanisms for Adaptive Locomotion: Lessons Learned From Nature
Poramate Manoonpong, Jørgen Christian Larsen, and John Hallam
Tutorial on 23rd of August, 2016.
Animals can autonomously move around in complex unknown environments. They can also deal with environmental and morphological changes. Furthermore, their movements give the impression of elegance including versatility, energy-efficiency and adaptivity. Biological studies reveal that one of the key ingredients underlying these complex achievements is the exploitation of neural mechanisms, soft materials and biomechanical structures. While this recipe has been identified, transferring it to artificial systems to approach animals in their level of performance remains a grand challenge. To tackle this issue, our tutorial will bring together leading experts working in the domains of soft robotics, biomechanics, and neurobotics, to show how soft materials, biomechanical structures, and neural mechanisms can be exploited to achieve adaptive robot locomotion and discuss future directions to overcome this challenge.
- Understanding the basic concepts of locomotion generation
- In-depth knowledge of embodied AI, biomechanics, and neural control for locomotion
- Understanding how soft materials, biomechanical structures, and neural mechanisms can be developed and exploited for adaptive robot locomotion
- Applying these techniques to develop robots.
The full-day tutorial is composed of the following parts:
- Theory: Invited talks by experts from soft robotics, biomechanics, and neurorobotics.
- Practice: Building robots that move using the Locokit modular robot construction kit.
- Discussion and future steps.
- Poramate Manoonpong, Associate Professor of embodied AI and robotics of the section of Embodied Systems for Robotics and Learning of the University of Southern Denmark
- Jørgen Christian Larsen, Assistant Professor of embedded electronics and AI of the section of Embodied Systems for Robotics and Learning of the University of Southern Denmark
- John Hallam, Professor of AI of the section of Embodied Systems for Robotics and Learning and Director of the Centre for BioRobotics at the University of Southern Denmark
Their main research focuses on 1) developing modular bio-inspired robots and their modular neural mechanisms towards embodied locomotion systems with adaptivity, energy efficiency, and versatility, and 2) using robotic and AI techniques and knowledge engineering to model biological systems. They are involved in many research projects, in the domains of neural locomotion control, biomechanics, body compliance, and biorobotics, e.g., ChiRoPing – Chiroptera, Robots, Sonar (FP7-project), CILIA – Customised Intelligent Life-inspired Arrays (FET project in FP6), BioAdapt – Biomechanics and Adaptive Neural Control of Animal and Robot Locomotion (BCCNII-project), Locomorph – Robust Robot Locomotion and Movements through Morphology and Morphosis (FP7-project).
09:00-09:30 Exploiting Soft Materials: “Soft robotics and adaptive locomotion: What an octopus can teach us” (Cecilia Laschi & Marcello Calisti)
09:30-10:00 Exploiting Biomechanical Structures: “Towards a robot that can develop body and mind together” (Fumiya lida)
10:00-10:15 Coffee break
10:15-10:45 Exploiting Neural Mechanisms: “From neural dynamics and synaptic plasticity to adaptive locomotion: An embodied neural computation approach” (Poramate Manoonpong)
10:45-11:15 Locokit: A modular robot construction Kit for building locomotion systems (Jørgen Christian Larsen)
11:15-12:30 Lunch break
12:30-16:00 Using Locokit to build robots that move: Lessons learned from lectures
16:00-17:00 Competition, discussion, and close