摘要: |
Older drivers stop driving for a variety of reasons, including being overwhelmed by the workload of the primary driving task. Workload estimates for specific routes from an in-vehicle navigation system or Google maps could provide guidance. More fundamentally, validated workload predictions are needed to allow the comparison of studies that otherwise seem uncomparable, for example, differing in terms of the number of lanes or the amount of traffic. In an ongoing M-CASTL project, the SAVE-IT equation (which utilizes data that are automatically collected in driving studies) was used to predict subjective ratings of workload in a driving simulator. The equation predictions were correlated with the latest subject ratings, but were lower than before because of previously irresolvable technical problems with the anchor clips (showing low and high workload) used to ground the ratings (and for other reasons). In this project, the researchers will develop and evaluate improved anchor clips that show all traffic (not just the forward scene as before and shown below) to appear on a 2D display in the cab. Scenarios for anchor clips will be driven in the simulator and then the recorded images will be combined in a variety of ways. To determine the best configuration, 12-16 subjects will answer situation awareness questions concerning candidate anchor clip configurations. Subsequently, another 16 subjects will drive the simulator and rate the workload of modified scenarios from the previous experiment using the new anchor clips. To determine the effect of engagement (ratings while watching driving are higher than ratings while driving because the driver is in control), there will be one test block where each subject rates clips as a passenger. This experiment will take about 90 minutes/subject. The result of this project will be improved, validated anchor clips for a follow up, on-road experiment. |