Michael C. Newman

Human perceptual overestimation of whole body roll tilt in hypergravity

Hypergravity provides a unique environment to study human perception of orientation. We utilized a long-radius centrifuge to study perception of both static and dynamic whole body roll tilt in hypergravity, across a range of angles, frequencies, and net gravito-inertial levels (referred to as G levels). While studies of static tilt perception in hypergravity have been published, this is the first to measure dynamic tilt perception (i.e., with time-varying canal stimulation) in hypergravity using a continuous matching task. In complete darkness, subjects reported their orientation perception using a haptic task, whereby they attempted to align a hand-held bar with their perceived horizontal. Static roll tilt was overestimated in hypergravity, with more overestimation at larger angles and higher G levels, across the conditions tested (overestimated by ∼35% per additional G level, P < 0.001). As our primary contribution, we show that dynamic roll tilt was also consistently overestimated in hypergravity (P < 0.001) at all angles and frequencies tested, again with more overestimation at higher G levels. The overestimation was similar to that for static tilts at low angular velocities but decreased at higher angular velocities (P = 0.006), consistent with semicircular canal sensory integration. To match our findings, we propose a modification to a previous Observer-type canal-otolith interaction model. Specifically, our data were better modeled by including the hypothesis that the central nervous system treats otolith stimulation in the utricular plane differently than stimulation out of the utricular plane. This modified model was able to simulate quantitatively both the static and the dynamic roll tilt overestimation in hypergravity measured experimentally.

Modeling human perception of orientation in altered gravity.

Altered gravity environments, such as those experienced by astronauts, impact spatial orientation perception, and can lead to spatial disorientation and sensorimotor impairment. To more fully understand and quantify the impact of altered gravity on orientation perception, several mathematical models have been proposed. The utricular shear, tangent, and the idiotropic vector models aim to predict static perception of tilt in hyper-gravity. Predictions from these prior models are compared to the available data, but are found to systematically err from the perceptions experimentally observed. Alternatively, we propose a modified utricular shear model for static tilt perception in hyper-gravity. Previous dynamic models of vestibular function and orientation perception are limited to 1 G. Specifically, they fail to predict the characteristic overestimation of roll tilt observed in hyper-gravity environments. To address this, we have proposed a modification to a previous observer-type canal-otolith interaction model based upon the hypothesis that the central nervous system (CNS) treats otolith stimulation in the utricular plane differently than stimulation out of the utricular plane. Here we evaluate our modified utricular shear and modified observer models in four altered gravity motion paradigms: (a) static roll tilt in hyper-gravity, (b) static pitch tilt in hyper-gravity, (c) static roll tilt in hypo-gravity, and (d) static pitch tilt in hypo-gravity. The modified models match available data in each of the conditions considered. Our static modified utricular shear model and dynamic modified observer model may be used to help quantitatively predict astronaut perception of orientation in altered gravity environments.

Pilot control and stabilization of a rate-controlled vehicle in hyper-gravity

Astronauts experience multiple altered gravity environments during space missions, in which they need to maintain sufficient performance for vehicle control tasks such as planetary landing and vehicle docking. We aimed to use hyper-gravity as an altered gravity test-bed to study pilot manual control performance. A long-radius (7.6 m) centrifuge (NASTAR Centers ATFS-400) was utilized to produce hyper-gravity. Subjects (N=12) were tasked with trying to “keep the cab as upright as possible” using a rotational hand controller in response to a pseudo-random roll disturbance in the dark. The control law was “rate-control” such that stick deflection was proportional to commanded roll rate, similar to a lunar landing vehicle. During initial exposure to hyper-gravity, pilot performance showed significant degradations relative to the 1 G performance baseline, in terms of increases in the root mean square (RMS) error in roll tilt. On average in hyper-gravity, RMS scores increased by 25% over 1 G levels. Subjects also reported increases in workload in hyper-gravity. Performance significantly improved in hyper-gravity over time. After several minutes the performance in hyper-gravity returned to near the 1 G baseline performance level. This is likely to be a critical time period for planetary landings with fuel constrained vehicles. These control impairments may impact not only flight performance, but also vehicle and crew safety.