| A number of studies have established the role of familiarity as an important factor of driving for older drivers. Familiarity is known to govern route choice, and there is a need for familiarity to be included in choice models. But very limited work has been conducted in quantifying the degree to which familiarity influences route choice, under what conditions, and how robustly models can approximate familiarity. This dissertation attempts to answer the first 2 knowledge gaps in Chapter 3, where analyses conducted on the factors that influence route choice among older adults showed that a) familiar routes were preferred as they were perceived to be shorter, direct, and had minimal traffic; b) different factors of familiarity were involved in choosing a route; and c) familiarity was the most important factor in explaining route choice after baseline route choice behavior, accounting for 26 percent of the explained variance. Additionally, extensive literature review revealed that current models of route choice that included familiarity failed to capture the multi-criteria nature of familiarity. In Chapter 4, this dissertation attempts to develop an abstraction hierarchy framework for describing the multi-criteria nature of route familiarity, and establishing a mathematical framework that can be used to calculate a new measure of familiarity---estimated route familiarity. The final chapter discusses the applications of the estimated route familiarity measure and abstraction hierarchy framework for the personalization of driver support systems and vehicle algorithm design. |
