It might surprise you to know that I actually hear from many of you after these newsletters are published. Most are laudatory thanking me for publishing the newsletter. A few tell me that they stop what they are doing when it arrives in their mailbox (probably because it’s so unusual to get real mail). Others write to ask questions or tell me how wrong I am.
Occasionally, I will hear from someone with a heartfelt story in response to my opinion piece on the front of the newsletter. The most responses were to the article I wrote on my father’s passing when I told the story of how he insisted I become an engineer even though he had no understanding with his second-grade education of what that meant. What always surprises and pleases me is when a spouse writes to me. Did you guys know that your wives email me? Honestly, I do hear from wives of control engineers that tell me how much they enjoy this newsletter.
Recently, I heard from someone reacting very strongly to my article on the Digital Revolution. In that article (featured in our March 2020 newsletter) I described what is called the “revenge of the analog”, how we are analog humans trying to fit our round analog butts into square digital holes. This provoked a strong reaction from an oil patch integrator in the part of the US where you find such businesses. If you were under the impression that everyone is on board with Industry 4.0, this integrator is evidence that you’d be mistaken.
First, some background. This gentleman has over 30 years of experience of mostly working in the oil and gas industry, ethanol conversion plants, and other process kinds of industries. He’s seen vast changes in the industry moving from a time of little-to-no connectivity to vast remote connectivity. He’s participated in automating process plants so that they no longer need vast armies of instrument techs, operators, and process engineers. He knows of what he speaks.
His first point was very direct. Process plants, he points out are filled with a plethora of legacy systems from Honeywell, Yokogawa, Siemens, ABB, and many more. All these systems are functional, well-understood, and maintainable but not designed to provide much in the way of connectivity. Unlike in the discrete world, many use Foundation Fieldbus and other process protocols (no EtherNet/IP or PROFINET IO there).
He asks, “Is the Digital Transformation and Industry 4.0 so valuable, is there so much ROI for these systems that justifies taking the plant down, replacing these systems, bringing the plant back online, retraining operators, process engineers and instrument techs on them?” He argues that no one can really justify that kind of expense and that it just won’t happen.
His second point was outstanding. He argued that sensorizing a process isn’t cheap as many like to say. Quality sensors are expensive and it’s not the direct cost where all the cost lays. It’s the ongoing calibration of those sensors that costs money. Every additional sensor requires the staff of instrument techs to be trained, equipped, and scheduled to maintain, operate, and calibrate those sensors. That process shouldn’t be thought of as a throwaway as it is by many promoting Industry 4.0.
Next, he railed against the consultants and some of his own peers who, for the sake of their own financial well-being are promoting these technologies in the process industry.
He said in part, “I have always wanted to write this myself and actually post this, however, I realize that would be ostracized by many of my peers who have bought into this and many who are riding this wave for a paycheck. Between the ISA and many other organizations, the IIoT, IoT, and European “Industrie 4.0,” these acronyms have spawned a new generation of specialists. I have discovered that a high percentage of these “digital gurus” have no comprehension of how the automation industry for the various industry sectors truly functions. A high majority have never set one foot in a control room let alone go up to the third level of a crackling furnace…”
I thought his next point was very enlightening. He informed me that a process plant is like a living, breathing entity. You can’t just equip it with sensors, process models, optimizations, and the rest and walk away. It requires continuous and constant tuning. As the process evolves, the feedstocks (raw materials in the process industry) vary, the systems evolve, plant personnel improves or new people with a different operating philosophy arrive and the operation of the plant is adjusted such that the models used to control a process plant become obsolete.
The point that he is making is that process plants are living, systems that grow, change, and sometimes deteriorate. Just as there hasn’t been an AI system that could quickly solve the COVID-19 crisis (maybe IBM’s Newton should stop playing chess and get a real job), a process plant is a very sophisticated, almost living system that, to date, hasn’t been reduceable to an optimization model that is very effective. Possibly, with the advancements in AI and machine learning that will change, but right now, he believes it can’t be done well.
Finally, he made the point that I’ve made time and again. Most IoT projects continue to fail. The published statistics from Cisco and others show that only a small number of projects are successful and only a very few generate any real ROI. Yes, there are some outstanding successes, but many fewer than the vendors promoting these solutions will describe.
I offered his comments to some of the supporters of Industry 4.0 to get a counter opinion. They countered that open interoperability will take time. Both OPAF (Open Process Automation Forum) and NAMUR groups are working toward that interoperability. They also agreed that there were IIoT charlatans that tried to make a quick buck in manufacturing and quite a few of those projects with those companies failed. They tried to make the point that there are new, low maintenance, and low-cost sensors that can alleviate the maintenance requirement. Finally, the supporters indicated that new closed-loop systems could manage the changing feedstock situation.
Personally, I am on the fence but leaning to the position of the system integrator. I believe that we are on the path but that it is going to take time, a lot more time than we think, to get to interoperability and Industry 4.0.