An API for location aware computing

Abstract

Indoor, location-aware devices have the potential to provide extremely useful resource and building information to human users; for example, a location-aware PDA could potentially provide map information in real time. We have developed a complete, extensible, indoor location-aware API to allow clients to interact with building and resource data stored on a central server. We have designed our own data model, emphasizing a space's adjacency relationships to other spaces as the most important relational characteristic. Route finding is a fundamental algorithm for the exploration of such a data model; we investigated, implemented, and evaluated several such algorithms. Furthermore, we have implemented both a server and several clients, demonstrating the robustness and reliability of the API. The API's most complex client, the CampusMap application, runs on a commodity graphics workstation and allows a user to interact with a full, three-dimensional model of campus. CampusMap generates lowest-cost routes between any campus location by interacting with a Location Server, visualizing campus spaces, and instantly updating route results.