Development and empirical assessment of a model of situation awareness for multitasking with locomotion

Human locomotion has long been considered an overly practiced motor behavior. However, recent research has revealed a demand of locomotion on attentional resources, especially when performed during multitasking. Situation Awareness (SA), a cognitive construct critical to decision making and performance in complex tasks, has been shown to be important while multitasking with cognitive and physical workloads. No research has been conducted on the role of SA during locomotion with perturbations (e.g., slips and trips) and concurrent cognitive task performance (e.g., walking and talking on cell phone).; The primary objective of this research was to develop a model of SA for multitasking with locomotion and conduct an empirical study to assess the validity of the proposed model for explaining proactive gait control in response to locomotion hazards. To support the empirical work, a virtual reality locomotion interface (VRLI) was developed to present walkers with realistic virtual locomotion environments (VLE) similar to everyday locomotion activities. An initial version of the VRLI consisted of a computer controlled treadmill, a head mounted display (HMD), and a graphical workstation running the VLEs and controlling the treadmill, based on participant movement using motion tracking sensors. The VRLI setup was validated through a pilot study that compared overground walking with treadmill walking in a VLE. Results showed similarities in walking characteristics between the conditions. Based on the pilot study, further enhancements were made to the setup. These included using a rear projection screen with a stereo projector and light-shutter goggles and a new treadmill with an embedded force plate (under the treadmill belt) for collecting gait ground reaction forces (GRF) and center of pressure (COP) data.; Using the enhanced VRLI, an experiment was conducted to evaluate the utility of SA during locomotion and validate the proposed model of SA for proactive gait control for responding to locomotion hazards. In this experiment, the controlled variables included navigation aid type (NT), a priori knowledge (AK) and perturbation cueing (PC). NT consisted of two levels---map-based navigation (MBN) and instruction-based navigation (IBN) and was manipulated between-subjects. AK consisted of three levels, low, medium and high, and was also manipulated between-subjects. The AK manipulation involved controlling the initial exposure of the walker to the test VLE and hence controlled their mental model development on the task environment. The low AK group was trained with a low fidelity VLE while the medium AK and high AK groups were trained with a high-fidelity VLE, but only the latter group experienced a perturbation. The PC variable was manipulated within-subjects and it consisted of combinations of visual cueing and physical cueing of locomotion hazards forming four levels---visual only, physical only, visual plus physical and no cueing. Dependent variables measured included a battery of GRF and COP variables along with response accuracy to SA probes presented using a real-time probing technique. Twelve males and twelve females from the NCSU student population participated in the experiment and performed the navigation task following four different routes in the VLE.; Results revealed participant proactive preparation for locomotion hazards, as observed through significant changes in GRF and COP measures. Effects included the nature of cueing of the perturbation and prior exposure to a trial with a perturbation involving visual cueing. There was also complex interactions between NT, AK and PC that revealed greater participant proactive control during MBN with higher AK under visual plus physical cueing compared to IBN with lower AK under visual only cueing. SA accuracy under MBN was higher for probes requiring subjects to project VLE future states, as compared to IBN.; Analysis of correlations between SA performance and gait response measure...