Understanding the I-IoT edge

The edge is the junction point between the industrial world and the IoT world in the cloud. The edge can be decomposed into three macro components, as shown in the following diagram:

The Edge Gateway is the core and is responsible for forwarding the data from the site to the IoT Data Hub, whether this is on the cloud or not. The Edge Tools are utilities for configuration, log management, and patching the edge operating system, either from a remote or a local user interface. The Edge Computing is a new, upcoming component that uses data to perform an action at site level or to provide insights to headquarters.

So far, the Edge has been limited to collecting and forwarding data using the Edge Gateway to the I-IoT middleware, whether this is on the cloud or not. Recently, however, industrial companies have been able to turn data into actionable intelligence using the edge computing, which is available at the edge side. In 2017, Gartner declared the following:

While this statement might seem a bit controversial, it highlights the role that the edge has played over the past two years. Industrial companies, after an initial phase of absolute cloud-centricity, have realized that it is not always possible to do everything in a remote location. The reasons behind this are as follows:

• Data export: National regulations could prohibit the exportation of sensitive or high-risk data. For instance, Saudi Arabia applies very strict controls on exporting data related to oil and gas. China does not allow the exportation of data at all.
• Network bandwidth: The network bandwidth may not be sufficient to transmit all kinds of industrial data. For instance, high frequency data, such as the vibrations related to a piece of rotating equipment, needs to be collected on a frequency in a range of between 1 and 50 KHz.
• Network latency: Advanced process controls or analytics related to changes in data for profiling asset behavior in a small time window suffer from high and variable network latency. In general, asset optimizations are required to carry out corrective actions as fast as possible within a deterministic time interval.
• Data connectivity: Process optimizations or asset maintenance are required to carry out actions or trigger the company workflow for the replacement of components without an internet connection.