I didn’t come from this world of OT or the manufacturing ecosystem. My first 20 years of professional work, from 1990 to 2010, were in an IT context as a product manager, general manager and finally vice president at Microsoft. There I learned, early and often, that the safe space for Microsoft employees like me was in IT discussions with IT audiences—CIOs, database admins, network architects, developers, end users—and if anything related to manufacturing came up, the rule was to get out of the discussion as quickly as possible.
Manufacturing standards (ISA-88), data types (time series), regulated industry requirements (CFR 21 Part 11), analyst firms (ARC Advisory Group), audiences (AIChE) and vendors (GE, ABB, Siemens, etc.) had more than a different language, they occupied a whole parallel universe (thank you to a former Microsoft colleague for providing the metaphor).
Now, 10 years after leaving Microsoft, including the last seven as a founding partner and executive at a manufacturing analytics software company, Seeq, I still struggle with the separate worlds of IT and OT. How do I explain to my ex-colleagues why there's such a gap between the IT and the OT sides of a manufacturing organization? Everyone talks about IT/OT convergence as a goal, which makes it pretty clear the two are "verged," if that’s a real word, with their separation considered an impediment to business success.
Roots of IT/OT divergence
The reality for many IT employees is that their last connection with OT engineers and managers as people, friends or acquaintances was in their senior year of high school when they took AP physics, calculus or maybe chemistry together. Maybe they were on the same sports team or together in student government. Maybe they dated. But after graduation the IT-to-be students went off to get liberal arts, business and computer science degrees, and the OT engineers went to a different campus, if not a different university, to pursue real engineering degrees.
I say “real” and I mean it; I remember when the state of Texas sued Microsoft in the early 1990s for having employees with “system engineer” on their business cards, when the employees were in fact not degreed or registered professional engineers. If you meet Microsoft Technology Solution Professionals, they may not know where their title came from, but now you do. And, from that early division into separate university programs, the chasm then created has endured.
I’ve heard all the jokes: IT is short for “internal Taliban,” those latte-sipping, loafered employees at headquarters, whose chief concern is keeping an email server running. Meanwhile, OT engineers work in triple-D plants (dirty, dangerous and distant), in mandated sturdy shoes and hardhats, with DOS-based PCs (gasp!) running critical processes—Luddites all. Whatever the finger pointing, the siloed, separate efforts of two groups with different priorities, locations and educational background persists.
Cooperation required
If you appreciate the separation of these two worlds, you can understand how spreadsheets are still the most common analytics solution for OT engineers in a world of big data and IT innovations. OT engineers have to rely on their own effort for insights, and can’t wait days, weeks or infinity for someone from corporate IT to help. Andy Bane, CEO of Element Analytics, recently captured the OT/IT attitude divide succinctly: “Their (IT and OT) technologies, processes, aims and professional worldviews are sometimes at odds, even if they share ultimate goals for business efficiency, continuity and profitability.”
Before I get any hate mail from Microsoft employees, let me recognize how much they, and other IT-centric vendors, have changed in recent years. Microsoft now has a VP of manufacturing in its U.S. subsidiary, along with worldwide VPs for energy, manufacturing and healthcare verticals, and has been staffing up manufacturing expertise in their sales organization for years.
The only organization to outdo Microsoft in its enthusiasm for capturing manufacturing workloads has been Amazon Web Services (AWS), which now has literally hundreds of jobs openings for employees with manufacturing expertise. As one example, at their annual user conference (re:Invent), AWS CEO Andy Jassy had a section of his keynote dedicated to AWS success in manufacturing. No coincidence in any of this: the tremendous focus on the business implications of IIoT, also called digital transformation, will be the key to the next generation of successful organizations.
We can think about digital transformation in the context of three evolving themes. The first is the business requirements for the next generation of insights and improvements. OT has been running on Lean manufacturing, Kaizen and Six Sigma for decades. What’s next and how can IT innovation be tapped to enable the continuous improvement objectives of OT? And how can a new generation of employees, digital natives who had computers in their cribs, be enabled to lead this next wave of predictive and optimization improvements?
The second are the tired yet true implications of the COVID-19 pandemic. In 2020, the cloud enabled innovation access at an unparalleled speed due to flexible access to resources and services. From a vendor perspective, 2020 was the sound of windows getting broken as old assumptions, objections and cautions were thrown out in a desperate attempt to adjust to the new environment of remote employees, wild swings in customer demand (remember oil at less than $0/bbl?) and supply-chain disruption. Flexibility, resilience and agility—delivered by the cloud—will be lasting implications of the pandemic to ensure organizations can rapidly adapt to unexpected market conditions.
Finally, there is the data reality: ever more data is being created and needs a place to go for storage, advanced analytics and contextualization by other manufacturing and business data sources. Kevin Prouty, VP of manufacturing at analyst firm IDC, expects a tripling of process data created per plant by 2025, and an increasing percentage of that data will be “born on the cloud,” meaning data generated by new sensors will be piped directly to the cloud.
Shotgun wedding
Which brings us back to the beginning: IT/OT and the cloud. As we look ahead, the ironic matchmaker of IT/OT convergence will be the cloud, the common denominator in the three trends outlined above. Specifically, it will be cloud platforms offered by IT-centric vendors Microsoft and AWS. Google, Alibaba and IBM may play bit roles in this drama, but the two leading vendors have a huge head start and lion’s share of the business.
The result won’t be the world I grew up in with employees sneaking their own PCs (Dell instead of IBM!) into the company to get access to innovative software (Lotus 123!). Instead, this is the cloud innovation world, where cloud-based applications have to go through an IT governance process.
The result is there's no such thing as a small software as a service (SaaS) application deployment. If you want SaaS then you need to go through IT. This means OT will be, and I apologize for being the messenger, co-dependent with their IT counterparts for cloud access to the applications they need to succeed. This could be the next generation of MES, EAM and analytics, or it could be simply leveraging an IT data scientist’s algorithm to optimize outcomes.
As a result, as an OT manager, you can try to fight that reality, or take your IT counterparts to lunch the next time you visit headquarters. Because in the cloud economy, you’re in the same boat, and it’s been a long time since you went to prom together.