1. Internet of Things Propels the Networked Society

Ericsson Research Blog

Internet of Things Propels the Networked Society


Ericsson is known for having a vision of 50B connected devices by 2020.


This builds on the proposition that anything that can benefit from being connected will be connected which is the foundation for the Networked Society. At the time of this writing [June 2012], 50B seems perhaps even a conservative prediction.

The Networked Society embraces all stakeholders: people, businesses and society in general. Different stakeholders have different interests and drivers for adopting ICT solutions. For people, it is more about lifestyle, fun and “wants” rather than “needs.” Enterprises are exposed to an ever-increasing competitive business environment requiring cost reduction, branding and differentiation. From a society perspective, saving energy, sustainability, efficiency and safety are important drivers.

Figure: Drivers and foundation of the Networked Society

Figure: Drivers and foundation of the Networked Society

The underlying fundamental enabler that makes this happen is technology evolution. The key enabling technologies are ubiquitous connectivity, smart devices, and the ability to integrate smart objects in different applications. We are now at the meeting point in time where viable technologies are available at the same time as concrete needs from the different stakeholders have emerged.


The Networked Society builds on personal communications as well as communication embedded in real world objects or things, i.e. both M2M and the Internet of Things (IoT), the latter representing the bulk of future deployed devices. The things we are interested in are very diverse and range from industrial machines to vehicles, appliances, lights, and buildings. The things are not limited to tangible objects; smart places and environmental observations are very important for many applications.

The application space is very wide. Improvements in traffic safety and traffic management is one example. Transforming the electricity grid to a smart grid, driven by new requirements like energy efficiency, microgeneration, electrical vehicles, and consumer energy awareness is another. Agriculture, water management and environmental monitoring are other less technology intensive usage areas.

These applications are already being deployed today, but the focus is on single applications and most of the time are characterized by “one device – one application.” In some cases, even special networks are being built for single applications. We do not believe this will lead to sustainable business in the end.

How can we benefit from the ongoing development, yet allow a richer, more open architecture to emerge? Can we reuse what we are deploying? In order to do this, we have to open up or even break the current application silos.


Instead of deploying devices with a single purpose or application in mind, we should allow devices to serve multiple applications, and applications to employ multiple devices. We should also open up and reduce application development costs and time to market by moving away from proprietary and legacy technologies and solutions.

The proposition is to move to a horizontal system with a focus on reuse of common enablers, and a true transformation to using the benefits of IP and Web technologies all the way, even in the tiniest device. Connectivity, access to data, data representation, and processing and storage elements are important common capabilities in such a system.

This will allow a truly open market to develop and deploy the different solution components, allow commodity components to be used, and enable easier interconnection with existing applications and Internet services.

Figure: Moving from silos to an Internet of Things

Figure: Moving from silos to an Internet of Things


Needless to say, devices are instrumental for the Internet of Things. We are already witnessing the deployment of a range of different devices. However, this development is only in its infancy, and to get to a true mass market, several technical and commercial challenges have to be solved. Costs for developing and manufacturing the devices need to be further reduced. The availability and compatibility of the devices to different environments need to improve. The ease at which the devices can be deployed also has to improve.

These challenges relate, in part, to ongoing technology development (such as advances in microelectronics and sensors), agreements on standards, reaching economies of scale, and business ecosystems to produce the right equipment at the right price. But one key issue is that the market is currently quite fragmented. Each industry vertical has developed its own technical solutions without much regard to reuse and commonality. In particular, for many industries (such as building automation), the current solutions inherit much from past legacy networks whereas off-the-shelf Internet technologies would, in many cases, have provided a much more flexible and inexpensive platform. In addition, even in a single industry sector, the number of alternative solutions is large. For instance, in building control and automation, there is KNX, LonWorks, X.10, BACnet and ZigBee to name a few.

What is needed is an architecture that is based on IP, a common set of application tools, and a reasonable set of link layer solutions. We believe that we should start by putting IP into even the smallest devices. Today, IP can run in very constrained devices as well as in very constrained environments [1]. The industry is already on this track as demonstrated by momentum in product releases, standards, and industry alliances such as the IPSO Alliance.

From a commercial standpoint, it is also important to build on link layer communications that support multiple applications. Deployment of new IoT devices on existing networking infrastructures is a natural requirement.

Furthermore, we should turn to widely accepted development tools. Today, development is often done with proprietary tools. Going mainstream means that we can make use of the thousands of developers out there. To this end, open APIs are also important, and the prospect of AppStores for IoT devices is attractive.

A key concept is that of embedded web services. Embedded web services are the means to get the valuable data in and out of the devices, using the well-established technology that is widely used by many developers. It will also ease the integration to existing Internet services and Enterprise systems. Variants of the Web Services model suitable for the tiniest devices have already been defined. For instance, Constrained Application Part (CoAP) [2,5] employs the REST paradigm but employs a more lightweight solution that HTTP.

Figure: The Embedded IP Toolbox

Figure: The Embedded IP Toolbox

It is also necessary to make simple profiles of the sensor data and there are efforts in this direction from both the research community and in standardization. CoRE link formats [3] combined with SenML [4] is one example. Examples of more heavy profiles that are dedicated include ZigBee Smart Energy Profile 2.0 [6], which basically is a vertical application profile that does not differentiate between the data and the application in which it is intended to be used.

Appropriate cloud-based application enablement services are required to ease integration of IoT resources in applications. These include managed connectivity services, IoT device management and IoT resource management. IoT resource management includes discovery and directory services, data capture and integration as well as IoT data and event processing like storage and stream processing. It is important that applications can expose their information to others, discover what other resources exist, and control how their own information is distributed further and federated.

Figure: Capabilities of application integration of IoT resources

Figure: Capabilities of application integration of IoT resources


Our vision of the Networked Society is not just about technology. It is equally important to create an ecosystem of device vendors, application innovators, network operators, infrastructure vendors, cloud service providers, and others to create a feasible business model that does not require application builders to excel in every area.

Ericsson takes a holistic view on the Internet of Things by driving the vision, the mentioned technology evolution as well as engaging and driving the necessary ecosystem formation. We also provide key enabling solutions to make the Internet of Things happen, like managed connectivity services for IoT devices via our Device Connection Platform and turn-key systems integration activities towards different industry sector applications. The Ericsson approach is to ensure that all the necessary parts exist for the stakeholders and user to benefit from the Internet of Things.

1. “Interconnecting Smart Objects with IP,” A. Dunkels and JP Vasseur, Morgan Kaufmann/Elsevier 2010, ISBN 978-0-12-375165-2

2. “Constrained Application Protocol (CoAP),” Z. Shelby, K. Hartke, C. Bormann, B. Frank. Internet Draft draft-ietf-core-coap (Work In Progress), IETF, March 2012.

3. “CoRE Link Format,” Z. Shelby. Internet Draft draft-ietf-core-link-format (Work In Progress), IETF, January 2012.

4. “Media Types for Sensor Markup Language (SENML),” C. Jennings, Z. Shelby, and J. Arkko. Internet Draft draft-jennings-senml (Work In Progress), IETF; January 2012.

5. “Constrained RESTful Environments (CoRE) WG,” http://tools.ietf.org/wg/core.

6. “ZigBee Smart Energy Profile Specification,” Version 2.0, ZigBee Alliance, to be published

Jan Höller, Ericsson, jan.holler@ericsson.com
Jari Arkko, Ericsson, jari.arkko@ericsson.com