Neuronal Control Mechanisms Underlying Animal Locomotion that Cope with Physical Changes in the Body and Environment

This project will lay a foundation for understanding of biological intelligence, and for the development of engineered systems that can process information, make decisions, and act to achieve desired outcomes. There would be significant societal benefits if the speed and precision of modern machines were matched by flexibility and resilience in responding to unstructured situations -- including those arising from interactions with humans -- rather than relying on preprogrammed instructions.

The human brain stands as a model of such flexibility and resilience, and translating the brain's functionality to machines is a flagship challenge for engineering researchers. Due to the complexity of neuronal circuits, however, precise mathematical description of brain functions remains an elusive ambition. This project focuses on a smaller network of neurons within the spinal/nerve cord, called a central pattern generator, with the goal of uncovering the mechanisms underlying a more primitive form of intelligence that underlies animal locomotion.

The research outcome will contribute to advancing scientific knowledge of biological information processing and to establishing a technological basis for enabling intelligent machines for various applications including autonomous robots for exploration of unknown environments and therapeutic medical devices for treating neurological dysfunctions. The multi-disciplinary research will be integrated with training and education of diverse student populations.

The goal of the project is to pose hypotheses on the design architecture of neuronal control circuits that achieve robust and adaptive behaviors during locomotion, and provide supporting evidence through model-based analyses. An initial working hypothesis is that the neuronal network embeds an internal model of the body-environment dynamics and can be decomposed into a diffusively coupled rhythm generator and a feedforward compensator.

The method is based on mathematical modeling of leech swimming, and computational and theoretical analyses using dynamical systems theory. The leech provides one of the best research platforms, because its neuronal circuits are relatively simple and well-studied, yet provide sufficiently rich control functions of engineering relevance. General control principles will be revealed to explain how the neuronal controller maintains or modifies the body oscillation pattern through sensory feedback when environmental changes or neuromechanical failures occur.

The outcome of the research is expected to contribute to the foundation of a new control paradigm that integrates trajectory planning and regulation into a distributed network.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.