Uma arquitetura para o uso de sistemas de localização outdoor e indoor por meio de dispositivos móveis

Abstract

Today, various technologies that visualize information from geographic positions are available for different fields of application. This information is especially useful when used in buildings such as shopping malls, airports, hospitals, supermarkets and offices, as they allow for faster localization of people and points of interest. The accuracy of these technologies can range from several meters up to a few centimeters. Outdoors, systems that use GPS have already proved their suitability in everyday use. However, in specific areas, such as the so-called urban canyons and the interior of buildings, these systems are unreliable, or at worst, do not work. In an attempt to minimize the limitations found in the use of GPS, several approaches propose to determine the location of people indoors. These approaches, for the most part, make use of an additional infrastructure that provides optical, acoustic or radio frequency resources. In addition, for mobile devices, work using such approaches does not primarily discuss the architecture used for its implementation, which makes its experimentation difficult. By only using sensory hardware integrated in mobile devices and data structures that represent possible routes in internal environments, the present work proposes an architecture for the development of systems that allow the exchange of location information among the participants of a network of contacts. In addition, the system allows displaying in digital map and Augmented Reality the location of these people, regardless of the space where they are. To compensate for the incremental errors inherent to the estimated navigation systems, a time series treatment algorithm known as Dynamic Time Warping (DTW) is used. The suggested solution can be implemented for external, internal, or both, without the need for the installation of a supplementary infrastructure.