Glare and Adaptation Effects Caused by the Miner's Caplamp

Abstract

Nearly all underground mine workers use individual caplamps with an intense central beam and a relatively dim surround. The small central hotspot (about 5–10 degrees in diameter) is over 100 times brighter than the wide-angle surround, creating an adverse ratio of luminances in the visual scene. Glare and transient adaptation caused by the hotspot make it difficult to see small or low-contrast visual targets in dimly lighted areas outside the hotspot. As part of a research program to establish minimum luminances for safe and efficient performance of various underground mining jobs, it was necessary to determine how the glare and transient adaptation produced by the caplamp hotspot affect visual performance in low luminance tasks in the mesopic region (0.03–0.30 cd/m2). If minimum luminance levels were determined with subjects fully adapted to low levels of even illumination, unrealistically low thresholds could be obtained due to the absence of the glare and transient adaptation effects caused by the ubiquitous caplamp in most underground work environments. An initial series of pilot experiments confirmed the importance of including realistic glare and adaptation factors present in the mining environment when conducting controlled visual performance studies in the laboratory setting. Accordingly, a number of actual miner's caplamps were obtained for use in the Mine Illumination Research Facility established at Perceptronics, Inc., in Woodland Hills, California. All visual performance testing includes the caplamp hotspot as it would occur in the visual tasks performed by underground mine workers. Data from four subjects established that “time to see” measures were significantly elevated when typical caplamp glare and adaptation factors were present. Data is currently being obtained on a series of visually realistic full-scale simulations of mining tasks, using appropriate caplamp luminance factors in the scene; this data will be discussed in detail at the '82 Annual Meeting of the Human Factors Society. The data from an initial experiment comparing caplamp illumination and even illumination (no caplamp hotspot) is reported in this detailed abstract. Four subjects, 42, 37, 19, and 19 years of age, with 20/20 vision and no known visual anomalies served as trained observers in a visual “seeing time” experiment to compare (1) caplamp illumination having a small hotspot and dim surround, and (2) even illumination produced by a point source (set to provide the same level of target illuminance as the caplamp) 15 degrees from the center of the 5 degree hotspot. The target was a Landolt ring (broken ring with thickness 1/5th its outer diameter of 110mm) with a 22mm gap (gap = 36 minutes of arc from the subject's viewing position 2 meters away). The Landolt ring was relatively large but low in contrast (-14 percent), representing a fairly difficult visual task at the low level of the target background luminance (0.048 cd/m2). In a paradigm designed to represent the commonly occurring adverse visual effects of caplamp lighting, the subject fixated a 10° diameter caplamp hotspot (6 cd/m2, at a distance of 2 meters) for 10 sec, then upon a tone signal from a timing device, looked over at the Landolt ring target, which was set in one of six possible positions (gap at 12, 2, 4, 6, 8, or 10 o'clock). The target ring was mounted on a large disc of background material (illustration board) cut in a circle to preclude overall shape as a cue to gap position. The surfaces were covered with paper of neutral matte reflectance so as to produce a non-specular contrast ratio between background and ring (1.16:1). The subjects looked at the target ring upon hearing the tone, then pressed a button upon seeing the gap, viewed without the transient adaptation and glare produced by the caplamp hotspot. The geometric means for the caplamp condition vs. the even lighting condition were 3.8 sec and 1.4 sec, with a mean difference of 2.4 sec (standard deviation of the differences = 0.7257, t = 6.6, df = 3, significant at the 0.01 level). In both conditions, the Landolt ring target had a background luminance of 0.048 cd/m2, but in the caplamp condition, there was a 5° hotspot (23 cd/m2) only 10° from the edge of the target ring (producing significant glare), and the subject had fixated another caplamp hotspot (6 cd/m2) for 10 sec just prior to viewing the target. This paradigm simulates the common situation where the mine worker has been viewing targets in the hotspot just prior to looking for targets just outside the hotspot (the eyes scan more than the head-mounted caplamp does). The differences in seeing times could imply that low-contrast hazards may not be seen soon enough.