Indoor navigation with SVG

This paper presents a mobile guide for our research department. It presents information about our research department, like general information, running projects, employees, research topics and more. The most innovative feature is however the combination of several indoor positioning technologies and a SVG based map visualization to help visitors to find their way on our floor.

SVG offers numerous advantages for mobile navigation including small file size, high resolution zooming, and animation. With the advent of the SVG Mobile Profiles and the emergence of high-performance mobile players, it is now possible to develop appealing navigation systems for PDAs and mobile phones that take advantage of SVG. In this paper we will report about our experience with several mobile players (TinyLine, eSVG and BitFlash) and describe in details how we took advantage of the BitFlash SDK to integrate the SVG visualization into the complete application. Besides this work, we developed a useful tool that allows to generate, given a common semantic representation of a floor plan, SVG maps adapted to the different players and platforms

We opted for the BitFlash SDK because of its good performance, the almost complete support of the SVG Basic and Tiny profiles, and last but not least the easy integration with Windows CE native applications. The result is a fast map viewer displaying the current position of the user, along with information about the surrounding environment, like points of interest, objects, etc. By clicking on a point of interest, the associated information is displayed by the application in a new page. For the navigation, the user has to define his desired destination, by clicking on the map or by selecting an entry from a list of predefined destination. The shortest path from his current location to his destination is displayed as a blue line and refreshed as the user moves around.

The computation of the shortest path is also done on the client. A network of nodes and edges built over the map roughly describes the possible movements. A Java-based application has been developed to edit the graph more easily. An implementation of the A* algorithm uses the graph to compute the shortest path from a node to another.

We worked with radio frequency and infrared positioning technologies. The complete coverage of the floor plan was achieved using the Wireless LAN solution from Ekahau Inc, with an accuracy of about 2 meters. A specific room was equipped with an ultra wide band (UWB) system from Ubisense Ltd, with an accuracy of about 15 cm. The mobile guide connects to a middleware server that combines the positioning data coming from the different systems and forwards the most accurate data to the clients. Infrared beacons were placed next to points of interest and could be used both for positioning and for catching the user's attention.