Will 2021 finally bring together IT and OT?
January 04, 2021
It's hard to think about the convergence of IT and OT as a prediction for 2021 because the broader concept has been the topic of conversation for nearly 10 years.
It's hard to think about the convergence of Information Technology (IT) and Operational Technology (OT) as a prediction for 2021 because the broader concept has been the topic of conversation for nearly 10 years. But IT/OT convergence was never going to be a big bang event. Elements have always been possible - even now.
Regardless of which side of that convergence you might rest your hat, everyone in manufacturing has a story that characterizes the two sides as being anywhere from merely disjointed to openly at war. Many recount open animosity between two irreconcilable parties.
There have been endless blogs and white papers written on the subject, yet there is still work to do to make convergence, and the benefits that can be gained from bringing the two parties together, a reality. Will we see it finally happen in 2021?
I believe that part of the reason for the underachievement in bringing IT and OT together has been that IT and businesspeople have been looking at it as a technology problem as opposed to a business problem. In the world of IIoT technology platforms, edge computers, and middleware vendors, the discussion is too often about how an individual’s offering can, ‘ubiquitously and securely bring any and all the data from everywhere all the time.’ The technology approach to convergence is foreboding in at least two ways.
First, the notion that all data, every sensor value, and every data base record should be available in case you need it, is often a false narrative. The reality is that I can’t think of any entity that wants to bring all possible data together all of the time. It might be a worthy cause but it’s a little like wall-papering the walls of a giant room with the pages of the dictionary in case I need to know the definition of a word. The vast majority of the time most of the information is irrelevant and the space that it takes up actually is a barrier to getting specific data when I need it.
The second overwhelming aspect of the technology approach to convergence is associated with security. Protecting the computing and networked assets of any manufacturing enterprise is obviously important. Both sides show well-placed enthusiasm and an adamant requirement for unwavering compliance to rules and regulations. However, the respective threats at the IT and OT level are often distinctly different.
Adherence to an unimpeachable technical approach to the IT / OT duality can at best slow down progress toward a workable solution and at worst scuttle the entire opportunity. Better mutual understanding and accommodation for distinct requirements is essential to solving omnipresent security requirements. Modern tools at both the IT and the OT layers allow for the well-disciplined and trained individuals to design and implement secure systems. I unwaveringly agree with the security mandate but do think that it needs to be addressed with a multi-layered approach.
Inevitably, nearly all of the discussions concerning IT / OT convergence include a discussion of the need to change the culture. What is less often discussed is what is causing the need for a change in culture and what specifically should be done to bring the change about.
With all due respect to the technology stacks and the security requirements, I think that the key to bringing about the IT / OT convergence is by taking an application-centric approach. What is the job at hand? IT and OT do not exist purely as academic exercises. Any application must deliver functionality that makes a difference in terms of either internal efficiency or the delivery of value to a customer. Many of these applications overlap the historical domains of IT and OT with less than stellar connection between the two. A contributing factor to the IT / OT gap is the difference in the way these respective groups perceive certain aspects of any application.
Applications in manufacturing that are typically procured by the IT group tend to be corporate solutions that may span the entire enterprise. Business finances, ERP and PLM typically fall into this category.
The culture around centralized applications includes a strong push for standardization. This penchant for uniformity is deeply ingrained in the traditional IT department psyche. One of the basic cultural tenants of IT has been “uniformity equals efficiency.” A single approach simplifies maintenance and support. Over the years there have been IT edicts surrounding both hardware and software technologies. Positions such as, “We are a Microsoft house,” or “We only buy Dell laptops,” are common among legacy IT communities.
I can’t argue that there is a rational argument for uniformity in terms of the cost of support. It also makes perfect sense for certain applications that can be reasonably applied across multiple variations of the same kind of environments. For example, a good manufacturing ERP can be applied across corporate plants that make a range of products. So a firm that makes plastic sheet, ceramic disks, and glass items can use the same ERP to develop BOM and routings to support the variation. Both the application and the IT team benefit from a centralized approach. So there is absolutely synergy between IT tenants and applications that are deployed this way.
This centrality and uniformity starts to cause friction. However, when they are applied to operational applications. One size fits all mentality runs counter to the individual plants’ reality. The day-to-day operations are based on the job at hand. The activities needed to produce a product grow out of and up from the processes necessary in that manufacturing. Centralized solutions that have to be extensively tailored for each operational plant result in core capability that matches none of the respective plants. There is no efficiency from attempted uniformity to a singular operational solution when that solution is applied to a divergent set of plants.
Rather, IT / OT convergence should be considered with an application-first approach. There are aspects of the operational application that can be effectively applied across multiple different operational environments. For instance, the capture and reporting of progress against an order translates well across plants. Whether the plant is making specialty chemicals or discrete parts, there is an interest by the ERP planner in the progress against the order. Extending the ERP solution to facilitate this operational activity can both provide the centralized group (IT) real benefit and will be accepted by the plants (OT) because it requires minimal tailoring to an existing required activity. This approach will create mutual value and avoid the friction based on a centralized edict to use a common MES that does not match the individual plants reality.
Defining a finer thread of application-centric capability that benefits from the convergence can deliver tangible results without the daunting mega-convergence effort. This consideration is more than just point integration between one IT and one OT application. Mutual acknowledgement of the IT culture of centralization and the OT requirement for tailored operations can identify additional opportunities for convergence.