This is the first entry in a series on the “Industrial Internet,” focusing on the basic elements, legal issues and procurement implications, principally from the perspective of the end user. The term is used to distinguish industrial and critical infrastructure applications from consumer “Internet of Things” applications, but similar concepts apply.
The unifying characteristic is that information on attributes of physical objects (or the human body with regard to wearables and medical telemetry) is acquired by sensors that digitize, analyze (to varying degrees) and transmit this data. Based on software programmed rules, the sensors may issue commands to actuators to change or modify the operation of physical assets. Sometimes the data is simply displayed and stored locally. An important function of the Industrial Internet is that the data is almost always subject to more in-depth analysis.
In over-simplified terms, in the Industrial Internet information is acquired from physical assets (electric generators) or local environments (refrigerated trailers (“reefers”)), by sensors affixed to or embedded in physical assets to measure specific parameters such as vibrations, pressure or temperature. Sensors often consist of software, firmware and a CPU and are connected to an RF transceiver or to a fixed wireline network (local or wide area).
As digitized, the data from sensors (different sensors measure different physical attributes) are transmitted (via wireless or wireline connectivity) to a local gateway, collection point or node (“node’) that, based on programmed rules and the information received, may issue commands to actuators (switches or valves) to shut down or modify operation of the equipment, lower the temperature, adjust the humidity, or trigger alarms for management intervention. In time-critical applications, the sensors may communicate with other sensors to take specific action.
After initial processing and commands by sensors or nodes (“at the edge”), the data is conveyed (real-time or not) to a “backend” (data processing capability (cloud-based or not)) that may either issue commands to the actuators or perform more in-depth analysis or both. This analysis may suggest changes in the prognostics or other programmed rules in the sensors or nodes, in data sampling frequency, or in the maintenance, manufacture or operation of the physical assets.
Except in enclosed facilities (such as factories or electric substations), the sensors or the nodes are often connected by one or more wireless pathways. The wireless data are typically routed to a wireline Internet connection, a MPLS port or a private network on to the backend. Industrial Internet communications are typically encrypted. Advances in operating system software and miniaturization (to accommodate local processing and issuance of commands by the sensors), IP connectivity, data management software, and “big data” processing capabilities enable the Industrial Internet.
The term “Industrial Internet” is something of a misnomer. An entity’s physical assets, its use of sensors (and nodes), and encrypted connectivity to the backend are typically a company-specific operation, not intended to be widely accessible. Thus, these networks may better be referred to as “Industrial Intranets.”
Footnote: This series is not focused on computer-controlled equipment, processes or technologies, such as robotics, used to produce refined products and chemicals, industrial equipment and consumer goods, collectively referred to as Industrial Control System (“ICS”) technologies. Auto assembly plants, refineries and soft drink bottling plants utilize ICS technologies.