Experimental and educational platforms for studying architecture and tradeoffs in human sensorimotor control

Abstract

This paper describes several surprisingly rich but simple demos and a new experimental platform for human sensorimotor control research and also controls education based on an off-the-shelf gaming platform. The platform safely simulates a canonical sensorimotor task of riding a mountain bike down a steep, twisting, bumpy trail using a standard display and inexpensive gaming steering wheel with a force feedback motor. We use the platform to verify our theory, presented in a companion paper, about how component hardware speed-accuracy tradeoffs (SATs) in control loops impose corresponding SATs at the system level, but also how effective architectures mitigate the deleterious impact of hardware SATs through layering and "diversity sweet spots (DSSs). Specifically, we measure the impacts on system control performance of disturbances and delays, quantization, and noise in feedback loops, both within the subjects nervous system and added externally via software in the platform. This provides a remarkably rich test of the theory, which is consistent with all preliminary data. Moreover, as the theory predicted, subjects effectively multiplex specific higher layer planning/tracking of the trail using vision with lower layer rejection of unseen bump disturbances using reflexes. In contrast, humans multitask badly on tasks that do not naturally distribute across layers (e.g. texting and driving). The platform is cheap and easy to build and use, and flexible to program for both research and education, yet highlights crucial gaps in both neuroscience and control theory that our new theory closes.

Internal Feedback in Biological Control: Architectures and Examples

“…there is also simple theory and associated experiments [10], [14]. Naively, success in the biking task seems to require speed and accuracy that the raw hardware lacks, making non-layered solutions infeasible.…”

“…The layered nervous system breaks the overall biking problem into a high trails layer of slow but accurate vision with trail look-ahead for advanced warning, and a low bumps layer that uses fast but inaccurate muscle spindles and proprioception to sense and reject bump disturbances. The motor commands from these two control loops to the muscles simply add in the optimal case, as well as in experiments [10], [14]. Effective architectures (such as layering) create a DESS where diverse hardware enables a sweet spot that is both fast and accurate.…”