Despite the growing prominence and capabilities of self-driving cars, little is known about human interaction with such autonomous vehicles. I am studying human interaction with multiple levels of automation and navigation in a simple driving simulator to discover principles for interaction design that encourage effective partnership between people and automated vehicles. I have conducted two experiments so far. In the first, participants completed driving tasks based on two different factors: automated versus manual control and navigation in low-level versus high-level language, which were combined via a factorial design into four conditions for each participant in the form of different driving scenarios. Performance was measured by travel time, frequency of driving and navigational errors, and levels of user satisfaction. The second iteration had all automated driving, with participants merely supervising while I gave varying amounts of verbal input.