There are many reasons that the measurement of drivers' mental workload has great interest these days and will increasingly enjoy this status in the near future. Accidents are numerous, seemingly ineradicable, very costly, and largely attributable to the human factor. Human errors in the sense of imperfect perception, insufficient attention, and inadequate information processing are the major causes of the bulk of the accidents on the road (Smiley & Brookhuis, 1987; Treat et al., 1977). Although both low and high mental workloads are undoubtedly basic conditions for these errors, an exact relation between mental workload and accident causation is not easily established, or easily measured in practice. The measurement of drivers' mental workload offers opportunities and pitfalls, as illustrated in De Waard's model of driver performance, demands, and mental workload (De Waard, 1996). Although stability of primary measures of driving performance over time is what the drivers' goals are, the conditions are variable and sometimes strongly demanding and require effort in variable "amounts" that at times are beyond capacity. The accident proneness that follows such conditions is the (for the time being irrefutable) rationale for the measurement of drivers' mental workload.
Abstract