Maciver, Malcolm, PhD
Tech Inst. D157 Evanston
|Areas of Research||
Behavior Cognition Language, Motor Control, Movement & Rehabilitation, Systems Neuroscience
|NU Scholar Profile|
|Recent Publications on PubMed|
Simulation of the neural input resulting from the signal input due to a prey during a prey capture strike by a weakly electric fish. A colormap of the weakly electric discharge the fish uses to sense objects in its environment is shown in the background. The prey causes a change in the firing rate of sensory neurons, also indicated by a colormap.
One way to view an animal is that it represents an answer to the question: “What’s out there that’s relevant to my fitness, and how do I get there to acquire those resources?” There is both the informational, or sensory problem of detecting and assessing these resources, and the mechanical problem of moving to the resource, such as food or a mate. The research in my group is dedicated to understanding fundamental problems of how an animal’s biomechanics relates to the animal’s informational needs, particularly how to solve the problem of moving through space towards a target of interest while simultaneously increasing the quality of the information extracted from the biosensor arrays on the body surface. We are engaged in both a basic science understanding of these mechanisms, and applying the knowledge gained for the development of new technology, such as novel locomotor and sensory devices, where the integration of movement control and sensing is of key importance.
One recent result helps illustrate our research interests. Analysis of reconstructed electrosensory stimuli associated with prey capture behavior of weakly electric fish revealed an omnidirectional sensing volume surrounding the body. Analysis of the small time reachable volume, a concept from control theory, revealed a similarly shaped locomotor volume that reflects the unusual maneuverability and backwards-swimming capabilities of these animals. The locomotor volume is nearly congruent to the prey sensing volume when evaluated over a time interval corresponding to the sum of the sensorimotor delay time and the braking time. This strongly suggests that biomechanical capacity and neural function are deeply interconnected. More advanced vertebrates such as mammals and birds appear to have sensing volumes that far exceed the size of their small-time locomotor volume, which may enable the execution of sequences of strategic movements (such as stalking a target) versus reactive control strategies. The degree of overlap between sensing and locomotor volumes, and their relative magnitudes, is likely to have fundamental impact on the structure and function of associated neural control circuits. Analogous issues arise in the design of autonomous robots, where reactive controllers are used to control motion within the stopping distance of the robot, utilizing fast proximity sensors, while switching controllers or planners utilizing slower sensors and more complex analysis of sensory signals are used to control motion over larger temporal and spatial scales.
We recently developed a method for computationally estimating the locomotor volume (green) and sensing volume (blue) of weakly electric fish, and showed that these two volumes are approximately congruent. This demonstrates the interdependency of the animal’s mechanics and nervous system.