How network communications are driving the IIoT revolution
The properties of IIoT (Industrial Internet of Things) communications differ significantly from more familiar control protocols, such as EtherNet/IP, Profinet, and EtherCAT. So how can manufacturers evaluate the many means of communication at their disposal to take advantage of the IIoT?
The first step is to realize that solutions based on IIoT and Industry 4.0 require detailed data. IO-Link is a standardized sensor interface, which provides access to detailed sensor/actuator data from the lowest field level, meets this requirement. While digital switch sensors only provide individual bits, IO-Link provides access to detailed identification, diagnostics and parameter data of a sensor or actuator.
Next, note that most automation systems are based on a PLC that contains the application logic. For applications such as controlling a robot arm, sensor data must be extremely accurate and reliable. Within milliseconds, the PLC calculates output signals and transfers them to actuators, such as valves and motors. Control-based Ethernet protocols such as Profinet, EtherNet/IP and EtherCAT meet these requirements.
Although these protocols are primarily based on the Ethernet standard, specific properties have been modified to achieve the high degree of timing accuracy and fast millisecond cycle times required in industrial applications. The data carried by these protocols can only be exchanged using specific hardware, such as a PLC, and processed using specific software offered by automation manufacturers. But transferring data from the ground level to higher level systems, such as cloud-based systems, is a basic requirement for the IIoT.
OPC UA, MQTT and REST APIs
This is where IIoT communications such as OPC UA, MQTT, and REST APIs come in. These communication methods rely less on real-time millisecond data and more on the availability of end-to-end data on different systems from different manufacturers. IIoT is not about individual process data from sensors and actuators, but rather the overall picture. The most important process may be collecting condition data to prevent future failure or detecting correlations in process parameters that influence product quality. In these operations, real time is not an absolute priority.
OPC UA is a comprehensive framework with sophisticated security mechanisms. One of the main advantages of OPC UA is that no specific device description file is required. Each individual device has all the necessary data, such as its own data structure, in a human-readable and machine-readable format. OPC UA is particularly suitable for large IIoT projects, where devices from different manufacturers have to be combined, but the network can still be dimensioned accordingly.
While OPC UA largely operates using client/server connections, MQTT is based on the publish/subscribe mechanism (note: OPC UA now also supports publish/subscribe, although most installations systems use the traditional client/server method). In publish/subscribe, a publisher (data provider) provides its data to a central server (often called a broker) on the network. Subscribers (data consumers) can flexibly subscribe to all publisher data or to individually selected topics. Since there is no permanent connection between the publisher and each subscriber, the data overhead for MQTT is significantly lower. This makes MQTT particularly suitable for networks with limited availability or where information needs to be transmitted to multiple consumers at the same time.
An API (application programming interface) is a programming interface provided on a device. REST stands for Representational State Transfer and describes the API design requirements. A REST API allows customers to create their own applications based on device data, where basic conditions are defined without obligation. APIs for an organization’s own devices are standardized within that organization. This makes REST APIs particularly suitable for applications where many different devices from the same manufacturer are used.
The main difference between IIoT communications and control network protocols is that IIoT communication methods have been designed to enable seamless, end-to-end data flow from sensor to cloud, thus enabling the huge potential of Industry 4.0 and IIoT.