John Fryer

The first article in this series discussed the Edge and Edge computing span which a broad range of definitions and applications. What the Edge means to a company, or an individual supporting “the Edge” is very much determined by their location and perspective. This piece looks at why Edge computing has an important role to play as industry evolves in the age of COVID19.

There has never been a more relevant time for Enterprises to implement a digital transformation strategy. The advent of COVID19 has forced every business to examine how they can operate efficiently with a (more) remote, and perhaps smaller, workforce. For many small businesses in service-related industries this has tragically proved impossible with devastating consequences for many families. At the other end of the spectrum, “knowledge based” businesses have generally been able to adapt quite well. Since much of their business and business processes are based on the Internet and the Cloud, migration to full remote operation is a natural extension of existing processes.

Between these two extremes there is a whole spectrum of businesses and operations that have various levels of automation and Internet connectivity. Many of these businesses fit with the broad definition of the industrial sector. The needs of each business will vary greatly based on whether they are process oriented, discrete, or a hybrid of the two. Size and scale are also key factors. The diagram below illustrates a basic breakdown for anyone new to the industrial and related sectors.

COVID19 has brought several critical issues to light:

1)     The need for more remote operation and monitoring due to staffing considerations at industrial plant locations

2)     Increased flexibility and rapid adjustment within the supply chain. This is perhaps more applicable in discrete and hybrid industries as they adjust production to meet changing demands

3)     Workforce flexibility to accommodate reduced onsite staffing, absence due to illness/quarantine restrictions, retirement of older workers and/or reduction in workforce driven by the economic situation

None of these factors is something that can be accomplished overnight, and all these considerations are an integral part of a Digital Transformation strategy. What can be accomplished, and in what timeframe is very much a function of the maturity of an organization, something I will cover in subsequent articles.

The rest of this article will focus on how Edge computing can be used to address the issues above, but first is it important to understand the architecture that has been implemented in most industrial automation environments over the last 40 or years.

Virtually all industrial operations have some level of automation. These may exist on individual isolated, but intelligent machines – packaging would be a prime example. More often, complete production lines may be fully automated, simply requiring operator oversight. Programmable Logic Controllers and Programmable Automation Controllers (PLCs/PACs) are the workhorses for running control loops in these environments.

At a higher-level, Supervisory Control And Data Acquisition (SCADA) systems have been used extensively and may include running complete plants. Large plants, such as oil refineries, have operated on a much larger automated scale with Distributed Control Systems (DCSs), although one might argue that with modern technology the difference between a SCADA system and a DCS is somewhat blurred.

The Purdue model has been used for many years within the automation industry to describe the various layers of automation as seen below.

While it has served the industrial automation industry well, the Purdue Model is control-focused with a heavy emphasis on layers 0-2. It is only as Enterprise networking became ubiquitous that connectivity and integration at layer 3 has become increasingly adopted. Crossing the divide between the “manufacturing” zone an the “Enterprise” zone is still something viewed with great skepticism by many in the operational technology (OT) world, if for no other reason than the fear of implementing cyber-security. This is a critical issue that any company needs to address and it a key driver for OT/IT integration. Cyber-security is one of the main reasons so many plants remain “air-gapped”, or at least believe that they are.

Leaving cyber-security aside for now, it is at layers 2 and 3 of the Purdue model where changes need to be made. This is where Edge computing plays a critical role. Much of the Edge computing model shown in the diagram above directly maps to the expansion of functionality at the supervisory control and manufacturing operations layers. With many architectural models and maturity models being proposed across the industry, it is important that a direct relationship can be drawn between these models, otherwise confusion will arise which will slow down adoption.

Implementing or enhancing Edge computing capabilities at layers 2 and 3 of the Purdue model is where companies can get a rapid return in the post-COVID19 world. In the scale of industrial automation, Edge computing platforms can be very cost effective and non-disruptive to deploy. Using virtualization, organizations can keep existing control and automation systems while adding new capabilities to address needs that have arisen while social distancing and near plantwide shutdowns have been in effect. These include remote operation, monitoring and automated data collection for existing analytics purposes. It means replacing manual processes, such as recording readings on paper, inputting them into spreadsheets and emailing them to someone to perform some basic analytics. The automated collection of data also enables the implementation of current generation technologies, such as Asset Performance Management (APM) and Manufacturing Execution Systems (MES’s) (where applicable), which can enhance optimization of the supply chain and help to streamline maintenance.

Edge computing also opens to the door to additional IIoT, Industrie 4.0 and Smart manufacturing technologies. This may mean that some upgrades to PACs/PLCs is necessary, or that new sensors and connectivity may be required to some degree. Once previously manual processes have been automated, it becomes possible to further re-engineer production more quickly and efficiently to support increased flexibility and agility. Utilizing the data collected on Edge computing platforms can also enable more advanced maintenance procedures, moving from time-based maintenance to condition based or predictive maintenance. This can be allied with Augmented reality (AR) and Virtual Reality (VR) to utilize plant personnel for maintenance. AR/VR can also be used to train new operators without in-person help. Edge computing platforms can also be used to provide coordination between robotic applications that replace the need for close human contact.

The extent to which a company can take advantage of Edge computing very much depends on the internal capabilities they already have. It is unrealistic to expect a company that has some automated machines that are not coordinated to suddenly jump to an automated factory employing sensors, machine learning analytics and remote monitoring. In subsequent articles we will explore the key problems that Edge computing can solve, what types of capabilities a company can expect based on the level of automation they already have and what organizations need to do to move forward.