Manufacturing digital transformation

Introduction

If you’ve attended any industrial automation conference in the last two years, you’ll note that there is one and only one theme: “The Digital Transformation promises to transform industries worldwide to achieve fully optimized and predictable asset performance. Manufacturers need to proceed with warp speed implementing it.” There is seemingly no other perspective and no other point of view on Digital Transformation, Industry 4.0, Digitalization, Smart Manufacturing and all the other names for this impending transformation.

This article is going to examine four big questions that loom large over Digital Transformation and make some guesses as to its future. It’s not that the intent is suspect. Who could disagree with these objectives:

• Establishing wired, wireless and Single Pair Ethernet as the sole physical layer for factory communications.
• Establishing a single communications standard to support everything from sensor to cloud communications.
• Decoupling of hardware platforms from software applications such that applications can run anywhere on the factory floor.
• Evolving PLC programming into a high-level system design language supporting distributed information and control systems.
• Radically reshaping the IT and OT departments. Restructuring these groups such that IT can worry about networks, storage, resources, containers, load balancers and the like while OT can manage applications, processes and devices.
• Enabling machine learning and artificial intelligence at the edge such that data can be processed and responded to quickly.

It’s unquestionably true that a system that met these requirements would save huge sums of money if it were to come to pass. The ARC Advisory Group has estimated that the DCS services market alone is $20 to $30 billion dollars, and a substantial part of that – as much as $5 to $10 billion dollars – might be saved by implementing these changes. And that’s just in process automation.

Can Digital Transformation Meet these Requirements?

OPC UA (with all due respect to MQTT) is the heart of Digital Transformation. It is still, without a doubt, the dominant technology for moving factory floor data into the enterprise and the Cloud. The OPC UA Foundation continues to enhance its functionality, its reach and features to meet the challenges that manufacturers face every day.

The Foundation is now moving aggressively to develop a Field Level Communications (FLC) standard. In FLC, controllers, devices and Cloud systems would all be connected using the FLC standard, replacing EtherNet/IP, ProfiNet IO and other current technologies. This nascent effort looks to require sophisticated data modeling, automation component standardization and a time sensitive network (TSN) for Ethernet deterministic behavior.
The vision of the OPC UA Foundation and others supporting the Digital Transformation is attractive, but the road to achieving it isn’t smooth:

Pothole #1: The Digital Transformation uses sophisticated technology and complex, interconnected systems – especially for cybersecurity (drawing on the right). These sorts of systems are unworkable for all but the largest of manufacturers. Devices in these systems will be complex to maintain and eliminate lockdown as a useful tool. The ability of small manufacturers to maintain these systems is problematic.

Vastly Increasing the level and sophistication of manufacturing technology while the number,
experience and training of the workforce decreases isn’t practical.

Maintainability decreases as the sophistication of these systems increases.

Pothole #2: The viability of OPC UA is worrisome given the recent report from the German Federal Security Office. A majority of OPC UA vendors (70%) reported that they continue to run the original version of OPC UA, and most have no plans to upgrade their offerings. The OPC UA Foundation can continue to add features and sophistication, but if vendors don’t implement it, the long-term viability of the technology is suspect. It’s worrisome to think that manufacturers faced with devices supporting various versions of OPC UA with varying features may experience increasing integration costs – exactly the opposite of what OPC UA technology promised to offer.

Reliability is compromised when manufacturers don’t have the data they need because of incompatibilities among vendors.

Pothole #3: Digital Transformation is an expensive endeavor. Devices with the kind of functionality required for Digital Transformation will be more costly than the relatively unsophisticated devices of today. If the past is any indication of the future, expect all but the largest and most wealthy manufacturers to continue to equip their manufacturing systems with low-cost devices that are not Digital Transformation ready.

To date, manufacturers have not been willing to increase their capital budgets to purchase more expensive devices when less functional, less expensive devices are available.

Pothole #4: The inertia of current ecosystems works against Digital Transformation. Allen-Bradley is the market leader in North America with a truly massive ecosystem of partners, suppliers, distributors, integrators and end users that all have a common, vested interest in maintaining the current environment. AB has been at the pivot point of almost every turn in the manufacturing world. Not the first to market with a PLC, but they had the first widely used one. Buying ICOM – the first provider of programming tools on Windows – moved the automation industry to adopt windows-based PLC programming. They were the market leader that promoted DeviceNet, tipping the industry from direct connect IO to networking. Ethernet on the factory floor became accepted when they adopted EtherNet/IP.

Each of these tipping points changed the direction of manufacturing automation but also served to solidify and enhance Allen-Bradley’s and its partners’ overall business. None were an existential threat to it. It’s difficult to believe that somehow that entire ecosystem in the US and elsewhere is going to readily agree to open systems, open hardware and run-anywhere open software that threatens both current product lines and business models.

It’s truthfully hard to envision such a thing or a tipping point that would lead to this kind of transformation.

Where do we Go from Here?

For the Digital Transformation to be successful, the prerequisites are clear:

Manufacturers must believe that massive value is obtainable – manufacturers, especially smaller manufacturers, must come to believe that the benefits of digitalization – substantial savings, profits and efficiencies – can result from the pain of this kind of transformation.

Manufacturers must be willing to reinvent their organizations – the complete implementation of digitalization portends a massive restructuring and reorganization of the responsibilities, roles and processes of both IT and OT. IT would have to take on the role of providing and supporting networking resources to the lowest level of the automation hierarchy. OT would change its role to become a manager of applications that run on the network equipment provided by IT.

The tools must be available, well-understood and standardized – Digital Transformation cannot happen without the platforms, software and tools supporting it. Vendors would have to provide the open hardware systems, virtual containers and operating systems to power edge devices. Software vendors would have to provide the plethora of open application tools to build the applications. Certification standards would have to be in place to govern how these open systems interact and exchange I/O and data.

There must be a smooth path from here to there – Unlike other kinds of digital transformations we’ve seen in music, transportation, film and elsewhere, a digital transformation of manufacturing won’t happen as a step function. There must be a smooth path from the semi-closed ecosystems of today to the open systems of tomorrow.

None of this is impossible, but it’s unclear what the tipping point might be that makes this happen.

Significant Digitalization Is a Long Way Off

Despite the benefits of digitalization, it’s likely that the interlocking ecosystems of trade organizations, vendors and user organizations will slow the adoption of any technologies and processes that disrupt the prevailing order. There is too much money invested, the ecosystems are too integrated and organizations just won’t change fast enough for a real tipping point that leads to massive change.

A future where most manufacturing automation systems are running on open platforms, the majority of process and discrete automation is using Ethernet and control programming is app-based using network resources provided by IT looks to be far over the horizon. There are too many cultural, organizational and financial roadblocks that are going to prevent widespread digitalization in the near future. Instead, look for very gradual, almost glacial change as the current ecosystem players slowly transform to a digital manufacturing world.

Connecting Traditional Industrial Equipment with MQTT