Adaptive organic electronics for sensorimotor learning
In this research direction, organic neuromoprhic electronics are developed for on-chip learning in robotics. As a demontrator, a robot learns to navigate in a maze to the exit, by creating sensorimotor associations though on-line training on the organic neuromorphic circuit. Such demontrations represent a fundamental step in which, low power and easy-to-tune organic devices and circuits, can function as adaptive elements capable of forming associative links for autonomous learning. In this way, adaptability and intelligence can potentially be distributed and placed conformally in any part of a robot and especially at the edge where the sensorimotor processes occur. By integrating sensory, actuating and learning primitives in-materio and at the edge, intelligence can literally be incorporated in the fabric of agents and not as a discrete form. This is in sharp contrast to spatially centralized and (mechanically) rigid computational cores such as in the contemporary neuromorphic chips and the discrete sensory/motor modular blocks of robotics.
M. Kuo, Lego Hogwarts, Lego Krusty Krab, Lego Death Star. Sure. Fine. A Lego robot’s organic brain!, Yale Scientific Magazine (2022).
A. Gumyusenge, Polymer-based electronics that can learn to drive: that’s smart, Matter 5 (8), 2439-2442 (2022).
S. Bolakhe, Lego robot with an organic "brain” learns to navigate a maze, Scientific American (2022).
I. Krauhausen, D. Koutsouras, A. Mellianas, S. T. Keene, H. Ledanseur, K. Lieberth, A. Giovannitti, F. Torricelli, I. McCulloch, P. W. M. Blom, A. Salleo, Y. van de Burgt, P. Gkoupidenis, Organic neuromorphic electronics for sensorimotor integration and learning in robotics, Sci. Adv. 7, 50, abl5068 (2021).
I. Krauhausen, D. Koutsouras, A. Mellianas, S. T. Keene, H. Ledanseur, K. Lieberth, A. Giovannitti, F. Torricelli, I. McCulloch, P. W. M. Blom, A. Salleo, Y. van de Burgt, P. Gkoupidenis, Local sensorimotor control and learning in robotics with organic neuromorphic electronics, Proceedings of Neural Interfaces and Artificial Senses (NIAS), 10.29363/nanoge.nias.2021.023 (2021).