The evaluation of drivers responses to a multi-characteristic power assisted steering system

Abstract

A sample of fifty male and fifty female drivers took part in an:. experiment designed to evaluate a multi-characteristic power assisted steering system. Subjects drove a car fitted with the system for two one-hour periods on public roads and on two test-track sessions during which a number of driving performance variables including driving time and steering activity were recorded. Drivers completed a specially developed questionnaire after each road drive. A subsidiary task, which involved the visual monitoring of an illuminated display and verbal responses, was administered during the test-track sessions. Factor analysis and discriminant analysis were used to analyse data from the questionnaire, road drives and test- track sessions. Data were first factor analysed and the factors subsequently used as variables in the discriminant analyses. It was possible to discriminate between male and female drivers, and between groups of drivers allocated to the different power steering characteristics on the basis of the discriminant functions derived. Thus, reales were found to be more sensitive to the force feedback characteristics of the standard power steering than females, finding it difficult to judge the amount of effort required to steer the car and tending to 'over steer' under some circumstances. Males drove faster than the females on the Motorway with the standard power steering, however, more slowly than females in urban driving, and drove faster and more accurately than females on the test-track. On the basis of the differences observed between drivers allocated to the different power steering characteristics, criteria were developed which allowed the specification of that characteristic which could be considerec 'optimal' for ordinary drivers of both sexes. This character- istic, termed "Speed Proportional Feel", provides the driver with full power assistance at low speeds, but increasingly inhibits the operation of the power assistance as vehicle speeds rise, giving more steering 'feel' at high speeds. The test-track data were further analysed by means of the analysis of variance and analysis of covariance. The results of the analysis of variance indicated that the presence of the subsidiary task had affected drivers' performance on the test-track. Analysis of covariance was used to provide a statistical control for the effects of the subsidiary task on drivers' primary task performance and a significant learning effect was observed. No significant differences were found in the number of subsidiary task responses made by drivers allocated to different steering characteristics. A recommendation was made for further research into the observed differences between males' and females' driving speeds which, it was suggested, may be related to the types of accident in which males and females are typically involved. Further research into the level of artificial 'feel' favoured by male and female-drivers was also recommended on the basis of the finding that females appeared to respond more favour- ably to a lower level of 'feel' than males.