ABSTRACT
Most in-vehicle navigation systems can provide drivers with turn-by-turn speech instructions. Users not only expect their navigation system to guide them to unfamiliar destinations but also to provide real-time guidance based on traffic congestion. In order to avoid time delays, stress, and frustration, navigation systems need to provide accurate, effective, and efficient information. Various studies have investigated user interface considerations for better speech guidance. Only a few studies, however, considered the effect of road characteristics on speech guidance and fewer yet focused on complex interactions that carry a high risk for navigation errors. In this paper we describe and analyze speech guidance for a unique set of complex interactions, namely, elevated roads. We focus on the components of speech instructions provided in current systems (maneuver and wording) and other potential information content. We report a field study in which we examined two navigation systems to record turn-by-turn speech instructions on elevated roads in Shanghai: GM OnStar, a server-based built-in navigation system, and Eroda, a portable GPS device. This study represents the first step towards enhancing the comprehensibility of a turn-by-turn navigation guidance for complex interactions.
