At a fundamental level, the Internet of Things (IoT) is about taking a considerable expansion of device connectivity and using that scale to expand capabilities dramatically. We have a few billion connected devices today, most of which require a human to operate, authenticate and secure. As we look toward 2020 – when Gartner projects there will be more than 20 billion connected devices – and beyond, most devices will not require a human to operate. The major communication pattern in IoT will be device-to-device communications. This paradigm will unlock tremendous capabilities, but true interoperability – not custom wired or bolted on – will be necessary in order for it to be possible.
Today, the world continues to see IoT primarily in terms of connections. However, it’s the transformational value of IoT on business that really matters, and that in turn will be the driver of connectivity. It’s about the expansion of capability, not the expansion of connectivity.
Before we go any further, let’s define capability in this context. Capability is a kind of unit service or function that can be provided by devices, networks and systems. Multiple capabilities together, or with new capabilities, can provide more advanced capabilities or services – a kind of capability expansion. But back to business.
Currently, we’re seeing the first stirrings of IoT in specific industry segments, but the proprietary nature of the connections means that the value add is limited to improving capabilities that already exist. Where true value gets uncovered – and industries transformed – is when you’re able to take various processes, each designed for a single task, and combine them into a new service in a simple and effective way. A great example is Uber: location, mobile e-commerce and crowdsourcing capabilities, each designed for a specific use, combined into a new form to transform an industry. The IoT (of which Uber is a part) will follow the same example. It’s not hard to envision other “application mash-ups” of this kind in the IoT, made possible by truly interoperable applications, device platforms and carrier networks. The possibilities are practically infinite, but achieving it will take true interoperability at all network layers, a high degree of data integration, management & analytics, and significant advances in standardization.
In order to achieve interoperability, it is necessary to take care of interoperability issues at all network layers. For example, on the network layer, it is necessary to design standard protocol translation mechanisms so that different entities (IP-based or non-IP-based) can talk to each other without any difficulty. Similarly, at the transport or application protocol layer, nodes speaking Hypertext Transfer Protocol (HTTP) need to also easily exchange information with nodes talking Constrained Application Protocol (CoAP). Such efforts may achieve a certain degree of interoperability; however, more can and should be done towards true interoperability (e.g. service layer interworking and interoperability). This will enable a true “Internet of Things” as opposed to a proprietary “Intranet of Things”.
In addition to communication interoperability, data integration and sharing can also be an essential aspect of this evolution. In particular, it’s imperative to sort out how to share massive data among different applications and organizations, and how to fuse valuable information from different data sources. This may even largely depend on the semantics on the data. For example, two pieces of data may have different names but in fact refer to the same thing. These kinds of technology need to be well supported so that data can be easily integrated, and information can smoothly flow across different organizations.
In addition, services can also be annotated with semantic descriptions so that advanced service workflow can be achieved. Standardized data formats and well-defined APIs among different capabilities and services will enable true interoperability at the data and service level.
Now that our well-integrated data is flowing efficiently, we need to address how it’s all going to be managed. There are too many types of data management to individually address here, but one representative example is data storage.
To achieve true interoperability, data storage solutions may need to store heterogeneous data with different formats, which may reside in different servers and distributed places. This kind of distributed storage needs interoperable interactions and interfaces among storage servers, in addition to interoperability between the server and the IoT device.
There are different ways for achieving the interoperability between the server and the IoT device based on scenarios. For example, in a scenario where the server and the IoT device are from the same IoT domain or application, data-level interoperability is likely not an issue as they may use the same ontology and have the same understanding of the meaning of generated data; but the server and the IoT devices in this case may use different communication protocols and an IoT gateway or proxy will be needed in the middle to perform translation and eventually achieve interoperability between the server and the IoT device. Alternatively, in a scenario where the server and the IoT device are from different IoT silo applications, abstract and semantic techniques (e.g. ontology mapping) will be additionally needed to enable data-level interoperability since each silo application may use different ontology and describe the meaning of data differently.
Storage is merely one consideration. Data management may handle the privacy and security issues of massive data from various data sources. Interoperability is also required for providing privacy and security solutions for different data in a network where devices will be able to self-authenticate. In order to support secure and private data level interoperability, a data management system needs to be in place that is flexible and robust enough to store and manage massive amounts of data from a variety of sources.
These elements of IoT are both foundational and critical. It won’t be enough to simply add billions of devices and connect them. In order for IoT to achieve true interoperability and for business potential to grow, these billions of connections will have to be organized in new ways. In these new paradigms of design, massive expansions of capability will be unlocked, and vibrant new markets will be created.
Dr. Chonggang Wang is a technical staff member at InterDigital, a company that develops wireless technologies for mobile devices, networks, and services worldwide.
Filed Under: Infrastructure, IoT • IIoT • internet of things • Industry 4.0