Application of Scheduling Methods in Designing Multimodal In-Vehicle Systems

Multimodal in-vehicle systems (MIVS) may improve time-sharing performance of drivers. However, it is not always clear for designers of MIVS about how to select appropriate modalities and determine the optimal order of messages presented to a driver. To solve this problem, this paper proposes a general procedure to select several scheduling methods (e.g., Johnson's rule and non-identical parallel machine scheduling methods) and uses these scheduling methods to assign appropriate modalities and determine optimal order of messages presented to a driver. An empirical study of an example multimodal in-vehicle system was conducted and it validated the effectiveness of scheduling methods as a tool to improve driver performance and reduce driver workload. Further extensions of the current methodology and usage of this general procedure to select other scheduling methods are also discussed.