Former NeuCo co founder Peter Spinney speaks about OPTIMITIVE
Jul 27 2021
I co-founded NeuCo in 1998 spent 20 years helping grow it into a recognized market leader in the nascent market for AI-based process optimization for power generation, including two years as part of General Electric. It was a wonderful experience personally and professionally, as well as an opportunity to travel the world. I then succumbed to the temptation to develop my own consulting business, which was also fun and fulfilling.
While I missed the experience of being part of a small, growing state-of-the-art technology company, I felt lucky to have had it. What I didn’t expect is finding an opportunity to apply this experience with another company that not only had developed cutting edge AI optimization technology, but was applying it across the whole spectrum of process industries striving to simultaneously improve efficiency, reliability, and financial performance. And not only that, but a company founded and led by a true technological visionary, Javier Garcia Sedano.
Like NeuCo, OPTIMITIVE was founded on the premise that to maximize its value, optimization technology required a broad, flexible software platform that leveraged multiple methods to extract knowledge from data and apply it to both operational and financial objectives. OPTIMITIVE, however, recognized from its inception that the same operational challenges and constraints can be found and addressed through a common platform across the process industries — including not only power generation — but chemicals, refining, metals, and paper production. This small but extremely capable company has demonstrated the ability to optimize all the relevant processes across these industries.
I feel so fortunate and gratified to be part of this great company. While industrial process optimization is likely to become much more widespread in the future, presently it is still a relatively young industry and a fairly specialized niche. A former NeuCo colleague once said, “If I get any more specialized than this, I’m going to end up knowing everything about nothing!”
But the intricately interrelated forces of competitive pressures, environmental compliance obligations, and technological change have combined to make optimization increasingly more of a “must have” than a “nice-to-have”. There are some parallels here to the overall evolution of industrial automation. When I was busy helping start NeuCo two decades ago there were still a significant number of large electric power plants operating with analog control systems. And many of the plants upgrading to modern distributed control systems (DCS) were doing so largely because replacement parts were no longer available for their OEM analog control systems. Now a days virtually all power plants still operating have a DCS.
Which brings me to a related point, maybe best conveyed by anecdote. Early in NeuCo’s history I was attending a conference sponsored by EPRI, where the resident controls guru Cyrus Taft asked for a show of hands by attendees whose fossil-fired power plants had been retrofitted with a DCS. Nearly every attendee raised a hand. He then asked, “How many of you that raised your hand are using your DCS to exactly replicate the analog systems they replaced?” Virtually all the same hands were raised.
The point Cyrus was making is that the industry had moved to adopt modern distributed control systems because those were the only systems still available, but was not making any significant use of the more advanced automation made available by such systems. One overly simplistic but still very much true characterization of optimization is that it exploits the automation capabilities available but not typically utilized in modern digital control systems. But optimization does more than that; it also tethers operational decisions to multiple objectives; and manages tradeoffs between such objectives while obeying specified operational constraints.
While the points discussed above were made in the context of controlling fossil-fired electric power plants, they are equally applicable to other continuous process industries, whether controlled by a DCS, PLC’s, or a combination of the two. While these industries have historically made greater use of advanced control, the majority of existing applications are limited to single objectives, and are often univariate, linking single controls to one objective.
Tethering all controllable parameters to specified objectives while managing the tradeoffs between them and obeying applicable constraints is what process optimization is all about. The exact nature and quantity of the associated benefits varies across industries and the processes they depend on. But the benefits associated with such optimization – and the challenges associated with doing it properly – are the reasons I am so excited to be helping OPTIMITIVE with what it does so well.
Peter J. Spinney