Finding objects using UWB or BLE localization technology: A museum-like use case

Some location-aware use cases imply that a person desires to find an object relative to his position. This is the case of museum visits or any indoors/outdoors activity where the objects of interest are not geo-referenced or even can move with time. For these use cases, a user does not need an absolute localization, but a relative information consisting of the range to the object and the heading angle between that object and the user's moving direction. In this paper we study two different radio technologies for finding an object of interest: Ultra-Wide-Band (UWB) and Bluetooth Low Energy (BLE). Range and orientation estimation performance is studied when using just the distance or the Received-Signal-Strength (RSS) provided by UWB and BLE systems, respectively. Both approaches are combined with Pedestrian Dead-reckoning (PDR) estimation in order to analyze the benefits that PDR information provides. For completeness we compare the cases where only one tag is fixed to the object to locate (the simpler and more flexible case) with the more ideal case where several geo-referenced objects of interest, each one with a tag, jointly cooperate to improve the relative location to one of the objects of interest (in both cases the user carries a mobile phone with BLE 4.0 or UWB radio). The UWB ranging radio, not common in most smartphones, but very accurate, is used as our reference to define the best achievable performance goal, in order to compare with BLE RSS-based performance. We demonstrate that the common BLE low ranging-accuracy technology combined with smartphone-based PDR estimation is capable, after some initial user's walking, of finding with decent range and heading accuracy the objects of interest in a museum-like set-up.

I. Introduction

Some location-aware use cases imply that a person desires to find an object relative to their position. When the absolute object's position is known and the user's location is also known, using a GNSS receiver or any indoor localization approach [1]–[3], the problem is straight forward. However there are situations when the object to locate is not georeferenced or its position can change frequently. Additionally we can be in a site where no location system is available or accurate enough to guide the user to the object of interest. In these situations, if the user is only interested in approaching a specific object, then a relative infrastructure-less person-to-object localization is a desired technology. Typical use cases of these situations occur in museums, where a person is interested in finding a particular famous painting or sculpture within a big hall or building [4]. The museum objects of interest can change its position from to time to time when new exhibitions are deployed. The user also in outdoor spaces could want to find objects of interest in parks or exposition gardens. For these use cases, we as users do not need an absolute localization, but a relative information that can be the range to the object and the horizontal heading angle from our field of view perspective.