Loosely Coupled GNSS and UWB with INS Integration for Indoor/Outdoor Pedestrian Navigation

Abstract

The growth of location-based services (LBS) has increased rapidly in last years, mainly due to the possibility to exploit low-cost sensors installed in portable devices, such as smartphones and tablets. This work aims to show a low-cost multi-sensor platform developed by the authors in which an ultra-wideband (UWB) indoor positioning system is added to a classical global navigation satellite systems–inertial navigation system (GNSS-INS) integration, in order to acquire different synchronized data for further data fusion analysis in order to exploit seamless positioning. The data fusion is based on an extended Kalman filter (EKF) and on a geo-fencing approach which allows the navigation solution to be provided continuously. In particular, the proposed algorithm aims to solve a navigation task of a pedestrian user moving from an outdoor space to an indoor environment. The methodology and the system setup is presented with more details in the paper. The data acquired and the real-time positioning estimation are analysed in depth and compared with ground truth measurements. Particular attention is given to the UWB positioning system and its behaviour with respect to the environment. The proposed data fusion algorithm provides an overall horizontal and 3D accuracy of 35 cm and 45 cm, respectively, obtained considering 5 different measurement campaigns.

Improving the Accuracy of Decawave’s UWB MDEK1001 Location System by Gaining Access to Multiple Ranges

“…The improvement using the redundant robust estimation is about 50%, which is by far better than other approaches found in the literature that makes use of median or moving averages as in [13] that account for improvements in the range from 10 to 20%. Our positioning errors are also better than other UWB solutions using a limited number of UWB anchors but complemented with INS+GNSS, as in [17], where authors declare a horizontal error of 35 cm. Our results are closer to the performance reach using sophisticated machine learning (ML) algorithms like in [14], with errors of 0.1 m, but these results are only valid after a learning phase which depends on the particular scenario, and could not be consistent with a different or changing scenario.…”

“…The accuracy is 1 mm and the 3D position registration is continuous on a logfile at a 1 Hz rate. This method, which is similar to the one employed in [ 17 ], allow us to record a ground truth (GT) that is used later to make the performance location assessment. The only movement restrictions are that, during the experiment, an LOS condition should be guaranteed so the S6 head does not miss the retro and get lost; also the mast must be kept vertical to guarantee that the x - y position of the retroreflector is the same than for tags.…”

“…Other approaches search for the improvement of UWB position estimates by fusing with other complementary sensors, such as inertial measurements units (IMU) or GNSS [ 17 ]. Of course, these approaches are desirable and effective, but are out of the scope of this work, where we focus on the analysis of multi-ranging redundancy for independent (UWB-only) position estimation using a specific commercial device and their limitations to gain access to multiple ranges.…”

On Hedgehogs and Marvelous Minds

“…Their "Creator" system is aimed at hobbyists and will connect with common, low-cost Arduino or Raspberry Pi systems. Although this system does not, currently, seem to be able to output data in an NMEA stream the way the Marvelmind system does, experiments with coupling accurate GNSS data with Pozyx positioning data seem promising (Di Pietra et al 2020). Similarly, a recent article by Almeida-Warren and colleagues (2021) investigates the archaeological utility of a related kind of low-cost device-the DistoX2-which couples a laser rangefinder fitted with a compass and clinometer, and it offers an additional alternative technological solution for environments where total stations and GNSS are not practical.…”

Innovations in geomatics teaching during the COVID-19 emergency

“…As widely known, Geomatics is the discipline that deals with survey, monitoring, and analysis of the natural and built environments from the metric and thematic points of view. Geomatics aims to provide a correct knowledge of the territory and environment; therefore, it is required in many circumstances: various surveying techniques and methodologies can be applied for detailed surveying of the position of objects, like technological networks or vehicles (Marzocchi et al 2017;Cristodaro et al 2019;Gonzalez and Dabove 2019;Leppälä et al 2019); for the production of cartography of territorially extended areas; for the monitoring of structures, infrastructures, and/or the natural environment (Bovolenta et al 2017;Minghini et al 2017;Gagliolo et al 2018;Themistocleus and Danezis 2018;Gagliolo et al 2019;Gobbi et al 2019;Barrile et al 2020;Campos et al 2021;Ferrando et al 2021;Vaccaro et al 2021); but also for innovative applications (Ferrando et al 2018;Dabove et al 2019;Piazza et al 2019;Di Pietra et al 2020;Rossi et al 2020;Benvenuto et al 2021). The instruments vary from traditional topographic methodologies to GNSS (Global Navigation Satellite System) positioning and from photogrammetry, also using UAV (unmanned aerial vehicle), to laser scanner technology.…”