Unlocking neural complexity with a robotic key

Peter Stratton, Michael Hasselmo, Michael Milford

Implementing neural systems on robots has successfully elucidated many principles of neural sensorimotor processing in animals. Despite some understanding of how the brain processes perceptual input and generates direct motor output, we only poorly understand what happens in between - the neural processes that turn sensory input into coherent behavioural output through time. Brains exhibit complex emergent properties that cannot be understood by studying neural components in isolation. Complex brains have evolved for comprehending and interacting with complex environments in the real world. Studies utilising neural controllers on robots in the real world can exploit real-world complexity (without having to model it) while simultaneously offering an entirely observable, fully controllable experimental model. Complex brains evolved in order to comprehend and interact with complex environments in the real world. Despite significant progress in our understanding of perceptual representations in the brain, our understanding of how the brain carries out higher level processing remains largely superficial. This disconnect is understandable, since the direct mapping of sensory inputs to perceptual states is readily observed, while mappings between (unknown) stages of processing and intermediate neural states is not. We argue that testing theories of higher level neural processing on robots in the real world offers a clear path forward, since: 1. The complexity of the neural robotic controllers can be staged as necessary, avoiding the almost intractable complexity apparent in even the simplest current living nervous systems; 2. Robotic controller states are fully observable, avoiding the enormous technical challenge of recording from complete intact brains; and 3. Unlike computational modelling, the real world can stand for itself when using robots, avoiding the computational intractability of simulating the world at an arbitrary level of detail. We suggest that embracing the complex and often unpredictable closed-loop interactions between robotic neuro-controllers and the physical world will bring about deeper understanding of the role of complex brain function in the high-level processing of information and the control of behaviour. This article is protected by copyright. All rights reserved.