The ultimate industrial edge computing device senses and controls physical processes, runs local programs, and communicates with industrial controllers, factory operations, enterprise systems, and cloud applications. This is required for profitable and competitive real-time closed-loop manufacturing business operations.

Over the years, industrial automation architectures have been characterized by the use of distributed computing to improve performance, quality, reliability, availability, responsiveness, and reduce software maintenance costs, resulting in ever-increasing computational loads for field devices. The limiting factors at each step are the cost, durability and reliability of the technology. This has changed as business, consumer and Internet of Things (IoT) technologies become ubiquitous in everyday life and advances in low-cost communications. An everyday device for many people, the smartphone is a clear example of a rugged, powerful computer that integrates communication and display functions.


The primary value of edge computing is to execute applications close to physical production, enabling fast response times and capturing real-time data with extremely low latency. The integration of higher-level functionality directly into this new class of powerful field devices and industrial controllers, coupled with real-time transactional business systems, is reducing the need for industrial middleware software. Business systems are evolving faster than industrial systems to meet the requirements of business functions including supply chain, customer service, logistics and Internet commerce. Mid-level software and computers have served the purpose of buffering, synchronizing, translating, and refining sensor and controller information, but have also produced fragile systems with large numbers of mid-level computers, duplicate databases, complex configuration controls, and expensive and difficult-to-maintain software. Edge computing is computing performed at or near the physical location of a user or data source. Distributed functions at the edge include optimization, expert systems, and artificial intelligence with new types of devices.


●Edge gateway supporting legacy systems

Industrial edge gateways are typically ruggedized industrial computers running middleware software that connects to PLCs, drives, and other edge devices to contextualize and map information to data enterprise software and databases. Edge gateways are ideal for extending capital equipment investments with installed legacy controls and automation to provide edge computing capabilities.

●Edge industrial computing platform

Rugged edge computing platform provides gateway functionality, among many others, including distributed control, optimization, web server, OPC UA server and client, artificial intelligence (AI), REST API, image recognition, and cloud communication (AWS, AZURE, etc. ). Many integrated multi-user environments, such as Docker and Kubernetes, support the addition of user applications written in standard programming languages, including Python and JavaScript.

●Intelligent/Intelligent Field Edge Devices

Smart/smart field edge devices are a new class of smart field devices including smart sensors and actuators that communicate directly with controllers, enterprise and cloud applications. These devices integrate distributed control functions including optimization, web server, OPC UA server and client, REST API and cloud communication (AWS, AZURE, etc.). User-based initiatives are defining new architectures based on these concepts, including the NAMUR Open Architecture (NOA) and Open Process Automation Forum (OPAF) standards.


With the integration of the single-pair Ethernet standard 10BASE-T1, the industrial edge is entering mainstream computing and IoT, enabling cost-effective embedding of IP communications into end-field devices, including sensors and actuators. Ethernet-based networks supporting industrial control and automation take advantage of the advantages of mass-produced Ethernet infrastructure products, including lower hardware, software, and support costs. SPE is the way to unlock more information directly from sensors, actuators, drives, motor starters and other devices.

Single-pair Ethernet is the basis for the Advanced Physical Layer (APL), bringing Ethernet to field-level instrumentation in hazardous areas. Field-level Ethernet will make the digitalization of the process industry a reality with its universality and rapidity. Current and voltage will be limited to provide an intrinsically safe solution for Zones 0 and 1/Div 1. The main goal is to adopt proven technologies and options in the field of process automation.

The Evolution of Industrial Edge Computing Architectures

Created on:2022-11-16