Industrial communication moving networks to the edge

Edge computing can help to determine the status of device, machine, line, and even plant health, which can be used to identify opportunities to increase overall equipment effectiveness (OEE). 

By Steve Fales March 23, 2022
Image courtesy: Brett Sayles

On-premise computing is a valuable tool that can help enable digital transformation in automation and is a good complement to the cloud for transforming raw data and diagnostics into beneficial prognostics, leveraging machine learning models, or creating easily customizable key performance indicator (KPI) dashboards. Edge computing can help to determine the status of device, machine, line, and even plant health, which can then be used to identify opportunities to increase overall equipment effectiveness (OEE).

While the cloud will retain its position in improving business processes and the customer experience by allowing access to the latest software, artificial intelligence algorithms, and global data scientist expertise. However, the cost of transporting and hosting data remotely makes it imperative to determine whether an application will benefit sufficiently from the advantages of the cloud to be worth the added cost.

An application requiring individual product customization and constant product innovation in a highly differentiated market may be a good fit for cloud-based analysis. However, if an application stays consistent, requires absolute uptime, and the market demands price over innovation and product uniqueness, then the lower cost and increased reliability of the edge may be a better fit. A critical need, such as requiring the highest level of security, may also determine whether or not an edge solution is appropriate.

Aggregating critical diagnostics

EtherNet/IP can help in aggregating critical diagnostics for further analysis, to monitor and improve plant performance at the edge by making valuable information consistent across devices and easier to access. EtherNet/IP now has a standard network diagnostic assembly that creates a scaled architecture concept where network and device diagnostic data that was once spread out in many different locations is now combined together to be more easily accessible. This helps minimize user programming, application development complexity, and the need for changes to devices on the network.

The EtherNet/IP standard network diagnostic assembly combines ten different attributes of three different object classes into one location, which requires only one message per device instead of ten. This reduction minimizes traffic, helping ensure optimal performance of the control network. The standard network diagnostic assembly allows for more valuable data to be transported from devices to the edge for analysis and action to improve the production process. The standard network diagnostic assembly is the first profile-independent definition of assembly instances in CIP with data applicable to any device, which results in a higher level of data standardization.

Edge computing should become part of a holistic automation strategy to allow for additional visibility into production process performance and to provide an opportunity to glean new insights into how to decrease operating costs, improve output, reduce defects, and reduce unplanned downtime. Intelligence at the edge is especially valuable when uptime and speed are critical, the consequences of a security breach are unacceptable, or the cost of the cloud cannot overcome the additional benefits. Over time, more advances will be made to improve the intelligence of devices, allowing for processing power to be closer than ever before to the applications for which they are the most beneficial.

– This originally appeared on Control Engineering Europe’s website. Edited by Chris Vavra, web content manager, Control Engineering, CFE Media and Technology,

Author Bio: Steve Fales is director of marketing, ODVA.