Bringing Intelligence To The Machine With Edge Computing

Growth in the industrial sector has always been driven by the introduction of new operating paradigms, which have improved its operations. For example, in the late 1980s and early 1990s, productivity was improved by Toyota’s introduction of lean manufacturing concepts that are still used today. The new millennium has instead brought the innovation of Cloud Computing, while today, we have entered the wake of industrial Edge Computing, also called industrial perimeter technology.

Knowing how and where it is appropriate to process data from the IIoT world is now of fundamental importance, considering the estimates of McKinsey & Company, according to which companies will spend between 175 and 215 billion dollars on IIoT hardware by the end of. 2025 globally.

Industrial Edge

Industrial Edge Computing is the ability to process sensitive data in real-time, closer to their source, allowing a better information flow and a faster decision-making process. The Industrial Edge Application is an application that takes advantage of the integrated infrastructure of IT and OT. The Industrial Edge Gateway is a single device with upstream and downstream connectivity and Edge Computing capabilities. Upstream connectivity can be towards the Cloud or other company systems.

The downstream is towards devices such as plc, Rtu, sensors, and other intelligent tools that use industrial protocols such as Modbus, EtherNet / IP Dnp3, Ied61850, OPC UA and MQTT.

Given the importance of data from production departments and IIoT devices, industrial Edge gateways have assumed particular importance, devices connected to various elements in the production and having the task of aggregating data to send them to end-points such as PCs, devices mobile and, in general, any device that connects to the local network.

The integration of these gateways, connected and monitored, within the OT infrastructure represents an important challenge for the industrial realities that intend to undertake the transformation with a view to Industry 4.0. There are numerous trends in progress: from the increasing affirmation of the OPC UA communication standard to the fact that multi-cloud architectures are increasingly common together with hybrid solutions in which processing processes are possible both in the Cloud and at the Edge or, again, on-premise. In general, the choice of where and to what extent to store and process data requires careful planning to avoid additional costs due to inefficient architectures.

The Opportunities Of Edge Computing In The Industry

The ability to make decisions in real-time, integrate optimized data and manage complex analyzes enabled by the Edge can bring the quality of industrial processes to a new level: machines and plants are monitored in an accurate but highly flexible way. Crucial in this context is the possibility of having Edge platforms open to interconnection with practically any machinery, even of the legacy type, and therefore compatible with the most varied protocols. The possibility offered by Edge Computing to have an “always-on” ecosystem, thanks to which even non-intelligent equipment can transmit their data, can drastically reduce unscheduled downtime, thereby increasing time, productivity and operational capacity of older plants.

The ability of the Edge removes the barrier to integration between IT and OT of compatible systems and technologies; furthermore, by distributing the processing and storage functions along the perimeter, cyber-attacks have no chance of affecting the entire network. Ultimately, the proximity of computational resources to the data source decreases risks and guarantees a high level of cybersecurity. The Edge has three main functional components:

• Edge connectivity: connectivity to industrial systems/collection and normalization of data for immediate use;
• Edge intelligence: local computation intelligence;
• Edge orchestration: application orchestration (creation, implementation, management and updating).

The ideal strategy considers all three of these components, using the Edge for device connectivity, data collection, real-time analytics, application orchestration, integration with Cloud and enterprise systems, and the execution of some machine learning algorithms. Therefore, a modern Edge platform interposes itself in that functional space between industrial devices and advanced analysis in the Cloud.

Examples Of Effective

Edge adoption is in predictive maintenance (if an anomaly in the operating parameters of a machine is detected at the Edge, it becomes possible to intervene promptly), or in the implementation of more flexible production and optimization or, again, in quality control.