Hyperlocal Context to Facilitate an Internet of Things

ABSTRACT

The Internet of Things (IoT) holds the promise for computers to understand our world by means of technology. Today, computers observe and identify the world, but unless they interact to optimally share all relevant information, their understanding remains incomplete. We argue that location-based device discovery is a prerequisite for the required IoT interactions. Based on this, we introduce the concept of hyperlocal context, a combination of device identity, location and information resources, which represents a digital understanding of the world at a human scale. The considerations for the implementation of hyperlocal context as well as current and future applications are presented.

Keywords

Hyperlocal context; Internet of Things; Location-based discovery;

INTRODUCTION

In the words of Kevin Ashton, who coined the term Internet of Things (IoT), technologies will enable computers to observe, identify and understand the world without the limitations of human-entered data [1]. Today, sensors observe our world in countless ways, such as temperature, light, proximity and sound. People and objects are uniquely identified through technologies such as Radio-Frequency Identification (RFID) and image recognition. Such observation and identification occurs without human intervention. But do these technologies understand the world? At what level of understanding is there an IoT? It is difficult to quantify an understanding of the world. However, it is clear that the level of understanding is proportional to the level of observation, which includes fundamental concepts such as identity (who/what), location (where) and time (when). For this reason, the IoT hinges on the ability for computers to maximize their understanding of the world by interacting to collect and distribute all pertinent information.

DISCOVERABILITY AND PHYSICAL LOCATION FOR REAL-TIME INTERACTION

In this paper we use the term device to refer to computers which observe and identify the world. We argue that location-based device discovery is a prerequisite for IoT interactions. Based on this, we introduce the concept of hyperlocal context, which represents a digital understanding of the world at a human scale, and present its real-world applications.

In order for two devices to exchange information, otherwise known as Machine-to-Machine (M2M) communication, the devices must first be aware of one another’s existence. For instance, a client device may be programmed with the Uniform Resource Locator (URL) of the server to which it sends data. In an IoT free of human-entered data, devices must autonomously establish such awareness. In other words, IoT devices must be discoverable in order for interactions to take place.

Consider the example of independent devices collecting information about a given space. Is the space as large as a factory or as small as a room? Are the devices observing the same metrics but at different locations within? Are people and objects being identified and observed as they move about the space? Are the devices themselves mobile? These static and dynamic factors influence which devices will share an understanding of the world at any given instant.

The physical location of the devices will largely determine the extent of the overlap in their understanding of the world. Moreover, the mobility of people and objects being observed and identified throughout the space introduces real-time dynamics. Imagine that our space is a conference room with a few environmental sensors. When the room fills with people carrying smart devices, each of which is equally capable of sensing its environment, the potential collective understanding of the space is greatly increased.

IDENTIFICATION, LOCATION AND CONTEXT

Location-based device discovery is predicated on an ability to uniquely identify and locate devices. Here we provide a brief summary of the relevant considerations in both of these domains, and how together they extend to provide contextual awareness.

Unique Identification

The importance of identifying objects in order to interact with them and bridge the gap between the physical and virtual world has been described in [15]. Barcodes are an example of machine-readable identifiers. The Universal Product Code (UPC) is a common standardized version of the barcode. The Electronic Product Code (EPC) [6] is an industry standard that generalizes the UPC to provide a universal identifier for physical objects anywhere in the world, which can be stored on RFID chips. Active RFID devices use an internal power source to signal strength and the known location of the routers themselves. This approach has recently been extending using Bluetooth Low Energy (BLE) beacons, which are essentially active RFID devices attached to specific locations, facilitating indoor positioning by mobile devices [18].

Location

Systems that locate devices in space are typically referred to as real-time locating systems (RTLS) or indoor positioning systems (IPS). One approach to location involves devices in fixed locations identifying themselves to their surroundings. For example, the periodic, uniquely identifiable broadcasts of WiFi routers can be used by a nearby listening device to determine its own location based on, for instance, their signal strength. The opposite approach is for fixed infrastructure to listen for the wireless transmissions of uniquely identifiable devices in range. This involves determining the location of the transmitting device similar to the infrastructure, which has knowledge of its own location.

While both approaches enable a host of applications, their current state suffers from limitations which hinder the progress towards a sustainable IoT. Both tend to be implemented as disjoint networks for dedicated services in local niche applications and are usually pre-configured to serve a single purpose service in a local network domain, which impedes the sharing of information among multiple networks and/or technologies. Moreover, these are often based on proprietary technology which makes standardization efforts impractical.

Many challenges remain and we see a legitimate need for new device discovery mechanisms within a framework of real-time location [19, 2, 25, 20]. RFID and location technologies are together key enablers of the IoT by contributing to the contextual understanding of a physical space through the identity and location of the devices it contains. The addition of sensor data collected by these devices, as well as accessory information enriches this contextual awareness.

IMPLEMENTATION OF LOCATION-BASED DISCOVERY

Mechanisms for location-based discovery are manifold. Here we present one such implementation which leverages location-aware smartphones. We comment on its suitability for IoT devices before presenting an alternative infrastructure-based mechanism which addresses the main challenges we have identified so far.

Location-Aware Device Discovery

Ambient social networking applications such as Highlight [14] Sonar [22], and SocialRadar [21] allow users to discover one another based on the physical proximity of their smartphones. These applications periodically send the smartphone’s location to the provider’s server such that a list of all participants and their locations is centrally maintained in real-time. Based on this list, the provider may calculate proximity and push relevant content, encouraging the participants to connect and interact.

However, such a mechanism is an unlikely candidate for the IoT for two reasons. First, location-awareness adds a non-negligible energy and computing burden, which is out of reach for many resource-constrained IoT devices. Second, this discovery mechanism excludes the participants from competing applications. The IoT requires vendor-agnostic device discovery which transcends walled gardens.

Infrastructure-Based Location

We argue that the devices of the IoT need not be location-aware. Instead, the communication infrastructure may assume responsibility for locating the devices it connects. Consider the case of a wireless device which communicates with an infrastructure node. If the location of the node is known, the device may be estimated as having that same location. This imposes no additional requirements on the part of the device.

Infrastructure-based location has the benefit of consolidating device identity and location information at the level of the infrastructure node. Imagine that each node runs a lightweight server providing an Application Programming Interface (API) capable of listing all connected devices.

In summary, an infrastructure-based location system which consolidates the identity and location of all connected devices is ideal, especially for resource-constrained, wireless IoT devices. The reelyActive infrastructure approach is specifically designed for this purpose and acts as a gateway between the devices of the IoT and the IP network. When the collection of all discovered devices is complete, accessible and organized based on location, this provides the foundation for contextual understanding.

HYPERLOCAL CONTEXT

Hyperlocal context combines, for a given space and time, a collection of all relevant devices and their properties and observations, such that computers can construct a digital understanding of the physical world. This requires careful design. Here we present some of the most prominent considerations.

Device Compatibility

As long as devices expose a unique identifier when they connect to an infrastructure node, they may be included in a location-based collection. Hyperlocal context requires resources to be associated with each device. This is subject to constraints related to the unique identifier needing to be matched with its corresponding resource and that resource must be available and shared.

Participation, Privacy and Security

Participation in hyperlocal context should be opt-in for the device owner. Observations, identities and locations may contain sensitive information which would require secure transfer from one computer to another.

FUTURE APPLICATIONS WITH BLUETOOTH LOW ENERGY

Bluetooth Low Energy is a recent wireless communication technology being adopted by many devices including the latest smartphones [3]. In our opinion, this is a promising technology for the IoT as it allows for the unique identification of a heterogeneous mix of devices with location at a human scale thanks to an indoor range on the order of tens of meters. At reelyActive, we have developed a BLE node so that these devices may easily contribute to hyperlocal context.

The IoT will enable computers to understand our world and hyperlocal context facilitates the interactions required for them to collect and interpret all of the supporting information. We have shown that location-based device discovery lays the foundation for hyperlocal context, and have described successful real-world implementations. It is our intent to advance the proliferation of hyperlocal context through open collaborations, so that it may evolve as an integral element of the IoT.

REFERENCES

[1] Ashton, K., That Internet of Things Thing, (2009) [Online]. Available: http://www.rfidjournal.com/articles/view?4986 [2] Bessi M. and Bruni, L. A survey about context-aware middleware, Italy, (2009) [Online]. Available: http://home.deib.polimi.it/bessi/obj/hm_paper/ContextAwar [3] Bluetooth SIG, 2012 Annual Report, (2012), pp. 5. [Online]. Available: https://www.bluetooth.org/en-us/Documents/2012AnnualReportFinal.pdf [4] Dashbook. Dashbook [Online]. Available: http://dashbookapp.com/ [5] Dungen, J., Antonescu, T. and Genest, P.O. Towards a Simple, Versatile, Distributed Low-Power Wireless M2M Infrastructure. M2MCIP, Sydney, Aus. 2013 [6] GS1. Epcglobal website, (2012) [Online]. Available: http://www.gs1.org/epcglobal [7] EPCglobal Architecture Framework [Online]. Available: http://www.gs1.org/gsmp/kc/epcglobal/architecture [8] Evrythng Limited. Evrythng [Online]. Available: http://www.evrythng.com/ [9] Fournier, P-A. and Roy, J-F., A Smart Shirt for Monitoring Heart, Breathing and Other Activities. IEEE EMBS, Boston, Mass. 2011 [10] Honkavirta, V., Perala, T., Ali-Loytty, S. and Piche, R. A comparative survey of WLAN location fingerprinting methods, WPNC (2009), pp.243,251 [11] Hui L., Darabi, H., Banerjee, P., Jing L. Survey of Wireless Indoor Positioning Techniques and Systems. IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, 2007 [12] Jongwoo S., Sanchez Lopez, T. and Daeyoung K., The EPC Sensor Network for RFID and WSN Integration Infrastructure. PerCom Workshops ’07. 2007 [13] Loke, S.W., Context-aware artifacts: two development approaches. Pervasive Computing, IEEE , (2006), pp.48,53

ACKNOWLEDGMENT

The authors would like to thank everyone who has participated in both our hyperlocal context experiments and the Log in to Life experience. We value your feedback!