Context

The ubiquity of mobile phones has made it very easy to implement outdoor location tracking solutions, typically using the Global Positioning System (GPS). When it comes to indoor tracking, however, the poor quality of the GPS signal inside buildings has led many researchers to turn to other mobile phone sensing capabilities, e.g., sound or electromagnetic waves, to achieve the same goal. All these systems, although offering many advantages, come up against the essential question of building indoor maps. Indeed, few indoor location systems have been deployed at scale, mainly due to the difficulty of finding suitable indoor maps. The reason behind this is the fact that each building is unique and unlike outdoor landscape, it is very difficult to automatically infer the inside layout of buildings.

Objective

The goal of this project is to develop a solution for automatically building indoor maps. Here many techniques already been proposed, such as Simultaneous Localization and Mapping (SLAM), which consists of constructing a map while tracking a mobile device. This approach can be extended by using a Pedestrian Dead Reckoning (PDR) approach, which uses widely available sensors, such as accelerometer, magnetometer, gyroscope, and WIFI signal signature detection, to estimate the position of a mobile device, while constructing an indoor map of the building. By distributing this technique across multiple mobile devices, the time required to construct the indoor map can be significantly reduced and the positioning accuracy increased. For this project, the student will rely on PDR as a basic technique and explore the possibilities of using a technology such as Bluetooth to report the proximity between mobile devices, in order to collect data and to improve the overall accuracy of the indoor map being built.

Contact

alpha.diallo@unil.ch