Abstract
Twenty three drivers were interviewed with respect to their opinions about communicating with in-vehicle systems by speech. Participants were asked which in-vehicle systems they would like to be able to control by voice, the dialogue they would like to use to instruct such systems, and how they would like it to respond. They were read a scenario which asked them to imagine they were communicating with more complex systems. Finally, the interviews allowed participants to reflect on the concept of communicating with in-vehicle systems by voice. Participants felt that the in-car entertainment system would benefit most from voice control followed by the heating and ventilation system, lighting and in-car telephone. Factors influencing this choice included familiarity with the features, their frequency of use and acceptability of current manual controls. Of the commands suggested in Part One and Part Two, 93% were terse, curt explicit imperatives. In Part One 89% of the terse commands contained less than four words, while the majority (53%) contained two. Analysis of terse commands in indicated that commands could consist of four different components: functions, settings, actions and directional statements. 39% of terse commands contained a function and a setting, 16% contained a function and a directional statement and 13% contained only a setting. The availability of user syntax preferences in the design of in-vehicle speech interfaces is discussed.
Get full access to this article
View all access options for this article.
