My research is focused on the design of location-based systems for indoors. We are exploring the available technologies in order to obtain a good balance between accuracy, response time and scalability within this kind of systems.
During my thesis I was advised by Prof. Oscar Canovas Reverte and co-advised by Prof. Pedro E. Lopez-De-Teruel.
We are currently working on the analysis of sensors available on smartphones, since these are the source of information that we need in order to obtain context information useful to identify the place where users are moving around and hence to infer their positions. In that sense, we analyse the information extrated from the WiFi interface, the camera, the accelerometer and also the digital compass.
Our main aim is to design a generic and extensible architecture, able to be adapted to heterogenous environments and able to offer differents location-based services to a wide range of devices.
Additionally we pretend to carry out technical studies to identify patterns of user behavior in order to provide location services with added value. Finally, we will discuss the security issues regarding to privacy on this type of systems.
In this thesis we present the architecture of a localization system that uses data from multiple sensors available in commodity smartphones. Our main aim is to provide a solution able to support accurate location-based services, such as augmented reality applications, pursuing a good balance between accuracy and performance. We present the architecture which encompasses the overall system proposed. This architecture supports the design and development of services able to combine different types of sensors. Its layered structure and the specification of well-designed entities makes it able to support different system configurations.
First we focus on the analysis of location fingerprinting techniques based on IEEE 802.11, feasible to determine the position of a device with a few meters of estimation error. Several refinements, based on the integration of additional sensors such as the camera or the inertial, are introduced to improve the efficiency of our solution. Using scale invariant features extracted from images we provide a solution for scene recognition that clearly improves the reliability of our result. Moreover, we take a step forward in the image analysis by including visual structure from motion techniques. It allows us to run off-line 3D reconstructions of the environment, and applying image resection techniques, we are able to provide precise estimations of both the 3D position and rotation of the device, obtaining an accuracy around 15 centimeters of error.
Our multisensor solution works in two different stages. We first obtain a coarse-grained estimation based on WiFi signals, digital compass, and built-in accelerometer, making use of fingerprinting methods, probabilistic techniques, and motion estimators. Then, using the images captured by the camera, we carry out the image analysis focusing on the subset of the 3D model spatially delimited by the previously obtained coarse estimation. Because of the difficulties found to build accurate 3D models in large and repetitive environments, our proposal makes use of state-of-the-art IMU data processing techniques during the training phase, in order to reliably generate 3D representations of the targeted environment. This process solves typical scalability issues related to visually repetitive structures in large indoor scenarios. The fact of getting high accurate 3D representations of the testbed scenario improves the efficiency of camera resection techniques, reducing the estimation error to 5 centimeters, with response times below 250 milliseconds. The set of techniques presented supports a wide range of location-based applications, from those requiring a coarse estimation to those with high accuracy requirements.
LOCUM: Project based on the development of location services for indoors. Link
Jose Antonio López-Pastor, Antonio J. Ruiz-Ruiz, Alejandro Santos Martínez-Sala, José Luis Gómez-Tornero. Evaluation of an indoor positioning system for added-value services in a mall. In Proceedings of the International Conference on Indoor Positioning and Indoor Navigation, 2019.
Thor S. Prentow, Antonio J. Ruiz-Ruiz, Henrik Blunck, Allan Stisen and Mikkel B. Kjærgaard. Spatio-temporal Facility Utilization Analysis from Exhaustive WiFi Monitoring. Pervasive and Mobile Computing Journal. December 2014.
Antonio J. Ruiz-Ruiz, Henrik Blunck, Thor S. Prentow, Allan Stisen and Mikkel B. Kjærgaard. Methods to Analyse Large Sets of WiFi Traces to Inform Building Facility Planning. In IEEE International Conference on Pervasive Computing and Communications (PerCom), March 2014. Link.
Antonio J. Ruiz-Ruiz, Oscar Canovas, and Pedro E. Lopez-de-Teruel. Practical image-enhanced LBS for AR applications. In International Conference on Mobile and Ubiquitous Systems (MOBIQUITOUS), December 2013. Link.
Antonio J. Ruiz-Ruiz, Oscar Canovas, and Pedro E. Lopez-de-Teruel. A vision-enhanced multisensor LBS suitable for augmented reality applications. In Journal of Location Based Services (JLBS), July 2013. Link.
Antonio J. Ruiz-Ruiz, Pedro E. Lopez-de-Teruel and Oscar Canovas. A multisensor LBS using SIFT-based 3D models. In Proceedings of the 3th International Conference on Indoor Positioning and Indoor Navigation, 2012. Video demo (Low resolution) (HD resolution).
Antonio J. Ruiz-Ruiz, Oscar Canovas, and Pedro E. Lopez-de-Teruel. A multisensor architecture providing location-based services for smartphones. In Journal Mobile Networks and Applications, November 2012. Link.
Antonio J. Ruiz-Ruiz, Oscar Canovas, Ruben A. Rubio, and Pedro E. Lopez-de-Teruel. Using SIFT and WiFi signals to provide location-based services for smartphones. In Proceedings of the Eighth Annual International Conference on Mobile and Ubiquitous Systems (MOBIQUITOUS). December 2011. Copenhaguen (Denmark).
Antonio J. Ruiz-Ruiz, Oscar Canovas. Integrating probabilistic techniques for indoor localization of heterogeneous clients.. En Actas de 10ª Jornadas de Ingeniería Telemática (JITEL 2011). Septiembre 2011. Santander (Spain)
Iván Lequerica, Antonio J. Ruiz-Ruiz, Andres S. Garcia, and Antonio F. Gomez-Skarmeta. An IMS based Vehicular Service Platform. In IEEE Vehicular Technology Conference Fall, pages 15, Ottawa, October 2010