TARSIUS. PERFORMANCE AND ROBUSTNESS ENHANCEMENT OF INDOOR LOCATION SYSTEMS FOR ROBOTICS AND ASSISTED LIVING

DESCRIPTION

The large advances in sensorial systems, wireless communications and onboard computation with mobile devices have made possible the construction of ambient intelligence-based spaces. So is the importance of this research field, that Horizon 2020 and the Spanish research program “Retos” include it as one of their main topics. It is essential for a smart space to have the ability to detect the presence of agents (people, autonomous vehicles) and to provide local based services (LBS). Despite the large previous research, the current indoor local positioning systems (LPS) do not reach the availability, accuracy and robutness that satellite systems (GNSS) get for outdoor use. The partners of this project have a considerable expertise in indoor LPS, and in the particular case of this proposal they envisage two main advances: performance and robustness enhancement of the weaknesses of the current LPS; and their use in future applications of high demand. The consortium has identified four main objectives. The fist one addresses the fundamental research in ultrasonic and radiofrequency positioning technologies, covering the mathematical theory of the position estimation, new signal encoding schemes for positioning (DSSS techniques and filter banks-based modulators) and the application of evolutionary techniques for model tuning and direct inference of the position based on the measurements; always focusing on feasible methods to be used in mobile devices (adaptation and implementation of algorithms with dynamic reconfiguration). Furthermore, this objective deals with the exploration of new non-intrusive approaches (device-free) to the problem of people location, such as tomography, TDV cameras and the disaggregation of the electricity consumption. The second objective focuses on the practical aspects of the indoor location technology such as the improvement of the data fusion methodologies that provide a robust and flexible way to use the available sensory information in each circumstance: beacon-based absolute positioning (US, RF, IR), embedded sensors (PDR and odometry), opportunity signals and map environment information. It also pursues the improvement of aspects related to the deployment and maintenance of LPS, with new automatic calibration strategies, extending its scope to 3D localization of unmanned aerial vehicles, with emphasis on the choice of upgradable modular systems and optimized algorithms. The third major objective addresses the integration of positioning networks with other networks (smartgrids and body sensor networks- BSN). Apart from the position, additional signals related to the human activity can be obtained. The process of this information through new artificial intelligence algorithms allow infer knowledge about the performance and potential needs of the intelligent space users. The last objective relates to the scientific and technological challenge of practical demonstration, with the modular design of sensorial networks and easy deployment for its use with people and autonomous vehicles, and the development of a case study with the monitoring of elderly and diagnosis of sleep disorders (with support of clinicians) by using positioning data, BSN and disaggregation of energy, prioritizing non-intrusive systems.