Manage Learn to apply best practices and optimize your operations.

Emerging frameworks for cross-silo IoT data models

As far back as 2015, in one of the earliest macro-assessments of IoT, the consultancy McKinsey quantified the market opportunity in nine categories. They ranged from consumer homes to wide open spaces and included home automation, factory, personal monitoring — health and fitness — and smart city applications. Each category, of course, can contain many hundreds and thousands of individual applications.

When the IoT market began to take shape, the initial focus of solution providers was on connectivity. There is a raft of market studies on the billions of potentially new connected devices for technology providers and connectivity service businesses to target. However, the work involved in connecting physical and virtual things to the internet is just a start. An increasingly important objective is to handle the data that comes from connected assets. That, in turn, means finding ways to connect, manage, analyze and, eventually, share data among different organizations.

In addition to the novelty of large-scale and distributed data management, the characteristics of IoT as a heterogeneous system of systems introduce an additional layer of complexity. Industry attention is now shifting from connectivity to new approaches to understand and interpret data. Heterogeneity introduces other new challenges, but there are issues in making data interoperable. This may be across application silos, between partners in a supply chain or between vendors of interchangeable devices and sensors.

The value of data models

A good data model can solve these issues. In other words, IoT data modeling offers an approach which could more efficiently describe, interpret, analyze and share data among heterogeneous IoT applications and devices.

Fortunately, several data models already exist. Many of them have been developed by different standards development organizations. Some of them are for specific IoT vertical applications or domains. For example, Smart Appliances REFerence provides a shared model for home appliances. Data models from the Open Geospatial Consortium are more for geosciences and environment domains. The Open Connectivity Foundation specifies data models based on vertical industries such as automotive, healthcare, industrial and the smart home. The World Wide Web Consortium Thing Description provides some vocabularies to describe physical things, but does not have much focus on data.

In contrast, operates as a collaborative community and it aims to provide more general and broad vocabularies and schemas for structured data on the internet. A collaborative approach to integrate and unify various data models is critical and necessary to work across IoT application and organizational boundaries. It requires cooperative efforts among different industry bodies.

OneM2M’s role in data model standardization

OneM2M, the global standardization body, was established to create a body of maximally reusable standards to enable IoT applications across different verticals. OneM2M focuses on common services, such as device management and security, that are needed in all IoT applications. The oneM2M standard takes the role of a horizontal IoT service layer between applications and the underlying connectivity networks. In doing so, it masks technology complexity for application developers and hardware providers.

Amongst its various standardization efforts, oneM2M takes a collaborative approach in developing its data model-related specifications. For example, one of its technical specifications, TS-0005 for management data model, is the result of a collaboration between oneM2M and the Open Mobile Alliance.

Another specification, TS-0023, laid the groundwork for a Smart Devices Template that was first applied to create a Home Appliances Information Model and Mapping. In the next release of oneM2M, Release 4, the underlying data model principles will be extended to support information modeling and mapping for vertical industries, such as smart cities.

New ideas for IoT data models

The focus on frameworks to manage data models across application verticals and domains sets the stage for the next phase of the IoT industry’s growth. This topic will be featured on the agenda at the forthcoming Internet Engineering Task Force (IETF) 104 Meeting on March 23- 29, 2019 in Prague, Czech Republic. Two IETF working groups, Constrained RESTufl Environments (CoRE) WG and Thing-to-Thing Research Group (T2TRG) will explore new ideas around IoT data models as they affect IoT application layer issues.

As one of oneM2M’s representatives at the IETF meeting, it will be interesting to hear about the latest progress on the CoRE and T2TRG activities. The meeting will also be an opportunity to explore new developments and possible collaboration opportunities around the use of data models for IoT management and security.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.

Data Center
Data Management