The Robot Vibrissal System: Understanding Mammalian Sensorimotor Co-ordination Through Biomimetics

Abstract

SummaryWe consider the problem of sensorimotor co-ordination in mammals through the lens of vibrissal touch, and via the methodology of embodied computational neuroscienceÑusing biomimetic robots to synthesize and investigate models of mammalian brain architecture. The chapter focuses on five major brain sub-systems and their likely role in vibrissal system functionÑsuperior colliculus, basal ganglia, somatosensory cortex, cerebellum, and hippocampus. With respect to each of these we demonstrate how embodied modelling has helped elucidate their likely function in the brain of awake behaving animals. We also demonstrate how the appropriate co-ordination of these sub-systems, with a model of brain architecture, can give rise to integrated behaviour in a life-like whiskered robot.

MIRO: A Robot “Mammal” with a Biomimetic Brain-Based Control System

“…A model of the basal ganglia (BG) [4,22] is then used to select, with persistence and pre-emption, one of these plans for execution by the motor plant at any one time. Overall, this system corresponds closely to similar, hardwired, behavior systems that have been identified in several animal species, including rodents [31] and amphibians [14].…”

“…Any behavioral plan is encoded as an open-loop trajectory for a point in the frame of reference of one of the robotÕs kinematic links. In all current plans, the point chosen corresponds to a Ògeneralized sensory foveaÓ [30,31] just in front of the nose; thus, MIRO is Òled by the noseÓ as a behavioral plan executes. The MPG comprises a kinematic self-model that is computed by moving the guided point and then identifying the remaining parameters (undriven) of the model through a principle of Òleast necessary movementÓ, starting with the most distal DOFs.…”

“…The lack of iteration limits the quality of the approximate solution, but is very cheap to compute, biologically plausible, and performs reasonably well. In previous work, we have used an adaptive filter model as a pre-processing stage to this MPG, greatly improving accuracy, and suggested that this may be a role played by mammalian cerebellum [31].…”

“…Therefore, some selection mechanism is required so that only one plan is disinhibited at any one time. There is substantial support for the hypothesis that the vertebrate basal ganglia is such a centralized selection mechanism, that may implement a form of optimal decision-making between competing actions, that operates across the different layers of the neuraxis, and that has contributed to the flexibility and scalability of the vertebrate brain architecture [4,10,22,31].…”

A future of living machines?: International trends and prospects in biomimetic and biohybrid systems

“…For instance, in Whiskerbot we looked at early sensory processing in an embodied spiking neuron model of the primary afferent nerve fibers coupled to fiber-glass whiskers instrumented with strain gauges and controlled by shape-memory-alloy artificial muscles. Our research with this model demonstrated to us the importance of fine control of the movement and positioning of the whisker for effective sensing [19] and led us to build additional degrees of freedom for whisker and head positioning into our later robots. As this work has progressed we have also added more elements to the brain-based control architecture.…”