Establishing a test methodology to evaluate the effects of sign parameters on the sign-processing task in a railroad environment

Signs located on either side of a railroad track provide different information including speed, location, and response (e.g., horn activation). The locomotive engineer sign processing task involves perception capabilities, demands divided attention, invokes working and long-term memory, is performed in a limited time capacity (directly related to speed of travel) and in a limited field of view. Complicating the task is a lack of sign standards.; One cause for a lack of sign standards is due to limited research regarding sign perception in the railroad operating environment. Conducting evaluations in a locomotive is cost prohibitive and can pose safety issues. A testing environment is needed to support measurement of sign processing in a manner that provides pictorial realism while maintaining experimental control.; This study proposes a testing methodology using high resolution digital images to replicate the constraints and provide sufficient detail of the railroad operating environment. An initial experiment (Experiment 1) evaluated whether digital images could be used to measure sign processing performance. Sign targets captured at measured increments in a field environment were presented on a high resolution monitor in a controlled laboratory setting. Sign processing performance in the field environment and indoor lab environment were measured. No significant differences were shown between field and lab results. It was concluded that digital images, as used, were adequate to evaluate sign parameters.; The second experiment (Experiment 2) applied the Experiment 1 approach to the locomotive domain by integrating important elements, compiled through research and other data collection efforts, of the railroad sign processing task. A rural railroad environment under daytime conditions was selected to measure the effects of sign location and font type. Individual components of the sign processing task were measured. Results showed significant differences for the factors of sign height, sign placement, sign shape, and font type. Follow-up pairwise comparisons and multiple regression analysis were also conducted. Based on these findings, a placement of 84'' opposite of the engineer's seated position with a middle and high test sign height (89'' to 136'') using a C-Series font were recommended. Future research of other environments and conditions is needed.