Item

Abstract

Purpose: To test the importance of geometric information (splay angle,
splay rate; Beall Loomis, Perception, 25: 481-494, 1996) and optical
flow information provided by road surface texture in steering
accuracy. Methods: Subjects drove along a simulated one-lane road
using a force-feedback steering wheel. The road was defined by either
(a) one continuous white line on a black background, (b) two continuous
white lines as kerbs, or (c) two lines and road surface texture. Turns
in the road appeared in random order. The subjects drove with constant
velocity of 16.9 m/s (60.8 km/h). Lateral deviation from the center
line, velocity and frequency content of the steering maneuvers served
as performance indices. Results: Most subjects reported finding the
task easier under condition (b) than under condition (a). Despite their
impression the data suggested different, counter-intuitive results.
Under condition (a) subjects performed more accurately (p < 0.01)
than under condition (b), and steering on a textured road (c) appeared
to be more accurate (p < 0.05) than on a road with no surface texture
(a). Conclusions: The difference between conditions (a) and
(b) may be due to the fact that apparent lateral shifts of the road
markings (splay angle) decrease with the distance from the road's
center line. Our results support the view that optical
flow obtained from road texture (c) enhances steering performance.
Currently we are testing increasing realism by shifting this paradigm
to a 180 deg projection screen.