Kevin Parker: What does it mean to say it's model-based?
By Kevin Parker, editorial director -- Manufacturing Business Technology, 3/1/2008 7:00:00 AM
At the recent ARC forum, Rockwell Software and digital manufacturing vendor Delmia jointly demonstrated a production-line simulation meant to support virtual commissioning of the real thing. In response to a question, those present were told the representation was a “logical” model, an “approximation of the physical world.”
Later that day, execs from process-industry vendor Aspen Technology talked about the company's use of a model, “a rigorous interpretation of the process,” to serve as the focal point in “design, operation, and maintenance” of production processes typical to chemicals, oil & gas, and other process industries. In this case, the model is necessarily more compute-intensive as it must mirror the physics involved in a way not needed in discrete manufacturing.
ERP provider Oracle was at the event as well, talking about the need for a business “model” where “the exception is the rule.” As is well known, the speed of advanced planning & scheduling systems follows from their “memory-based model” approach. And as sales & operations planning evolves into a kind of total business planning, it too uses a model-based approach to simulate alternative scenarios in support of sound decision making.
Finally, at Autodesk World Press Days, held the week following the ARC forum, much of the talk was about use of a 3D digital model to “visualize, simulate, and analyze” a product before it's built, and constituting a kind of digital prototype. The model includes “form, fit, and function,” i.e., the geometric definitions of the parts and their relationships, and increasingly, information on how they work together.
Despite being applied in fundamentally different realms, the model-based approach seems to be very important to the way computers will be used in manufacturing and supply chain. Models will also lead to fundamental changes in the way decisions are made—whether in the design and operation of a production line, a production process, a supply chain, or a common household product.
Chess provides the fundamental paradigm. It's only in the last several years that computers, by comprehensively crunching the numbers and weighing all possible moves, have been able to defeat even the grandest of chess grandmasters who, at some level, remain dependent on intuitive insight. Production managers, engineers, and designers should increasingly find themselves in the same boat.
This doesn't mean anyone is out of a job, especially in today's skills-constrained environment. And it should be noted that in chess circles serious consideration is given to the idea of tournaments that pit humans, each supported by a computer, against each other.
As use of models-based solutions grows, so too will dependency on them, as will the challenges incumbent in keeping the virtual environment synchronized with the “real” world. Vendors say that automated means for incorporating new “objects” into simulations are coming, as well as change management capabilities tied to work flows.
Another example, someone pointed out, is that through model development, a chemical engineer sees DCS screens that mirror the physics involved. But how do you make that meaningful for the operator? In other words, how do you shield the user from the model's complexity without making them become an expert in modeling?
