Peeling Away the Layers of IoT Interoperability
Interoperability issues are not new; but with more than 800 vendors, 300 platforms, and over a dozen standards groups involved, the Internet of Things (IoT) produces a wide range of possibilities and conflicts on wired and wireless networks.
The Layered Approach
An IoT interoperability analysis approach is to describe the layers of the hardware/software stack:
- The device layer is designed to seamlessly accommodate and integrate new and existing devices into the IoT ecosystem. The devices can be hand held or embedded in other devices like cars, trucks, trains, planes, construction equipment, and manufacturing systems. The number of IoT devices will eventually exceed the number of human beings on the planet, making asset management and identification a significant challenge (see " Are You New to IoT?").
- The networking layer supports object mobility and information routing independent of the information interacting with the applications. New wireless technologies will have to be part of the layered approach (see "LPWA Live for IoT").
- The middleware layer will be very important for seamless service discovery and management of smart endpoints. Many different hardware devices and networks may be involved in the process of collecting and uploading data to the cloud. We cannot expect all these endpoints to interoperate without some conversion.
- The application service layer's primary function is to introduce physical endpoints into the information world by reusing different parts with dissimilar elements or constituents (heterogeneous) application services from heterogeneous processing platforms.
- The data and semantics layer provides a common understanding of the IoT data and its analysis.
This layered approach is one way to analyze the IoT problems. There will be other approaches to defining the layers. If the layered approaches don't line up, then the standards, products, and services will not deliver a unified IoT approach. There will be industry, vendor, and service provider silos.
One trend in IoT is moving processing from cloud computing to the remote network known as edge computing (also known as fog computing). Machine-to-machine (M2M) communications has been characterized by the role that cloud platforms play in enabling processing applications to be run on the IoT data. It is now recognized that computing resources on the network edge offer faster processing than those in the cloud. Fog computing adds architecture choices and allows those deploying IoT to improve physical assets and processes in new ways. The variation of the IoT processing means that fog computing will have to be part of the IoT layered approach. With fog computing, all the layers may reside at the edge in combination with the cloud computing facility (see "IoT at the Network Edge").
Interoperability solutions will not work if there is not trust among cooperating entities. Trust management is a key issue that must be addressed when working with independent stakeholders. In addition to a common layered approach, there must be a trust management framework aimed at supporting interacting endpoints. Evaluate trustworthiness since many of the components and services will be from different sources.
Many in the IoT industry see governments as hindering the rapid development of IoT. But these parties do not have vested interest in the privacy, security, and safety of the IoT ecosystem except to avoid conflicts and damages. It is up to governments to mandate protections for how the data is measured by the collector, distributor, and processor, or else the analyzer could falsely assume that it is correct. But what if it's not? How can this be remedied? Customers need to spend time and money trying to correct their erroneous data, assuming they can find out what's incorrect in the first place.
We need to create policies and regulations to deal with IoT data collection processing and distribution. Consider a social network posting that is false. Will it ever be completely purged form the Internet?
About two years ago some analysts thought that the significant number of IoT standards would merge by 2017. Unfortunately, this has not been the case.
Some of the players in the IT community offer established technologies with layers of interoperability such as the ZigBee alliance, which calls for universal language for IoT. But there are still many choices for developers and consumers. This prevents IoT from being simple and easy to use. It could be as long as three years before we start to see a consolidation. Unfortunately, every computing and silicon vendor is competing to dominate the space, so right now it appears that we will have more of IoT silos than anything else.
My previous blogs on IoT include: