Training Approaches for Using Simulators to Teach Process Control Systems
Simulators are a powerful tool for training operators, but how you use them has a major effect on overall effectiveness. Don’t let technology make you forget the human factors. Insufficient perceptions and philosophies of adult learning and out-of-date training methodologies stubbornly persist. These can diminish the effectiveness of training, regardless of the sophistication of the simulators employed.
If you use simulators or you’re considering using them to train control room operators, it can be very helpful to consider different points of view and philosophies of adult learning to allow for more effective training program design. Training methodologies and activities are described here for both custom and generic simulators.
In the last 30 years, digital control systems have fundamentally changed the way work is accomplished in process industries. These remarkable machines have become supremely complex, inexorably automated, and increasingly challenging to learn.
We often attempt to teach people to operate these systems by employing powerful and innovative process simulators. To effectively use simulators we absolutely must employ equally innovative ways of training. However, insufficient perceptions and philosophies of adult learning and out-of-date training methodologies stubbornly persist. These can diminish the effectiveness of training, regardless of the sophistication of the simulators employed.
“How do we explain this poverty of training in the face of a massive technological expenditure? In part, it reflects a profound underestimation of the skill demands associated with a technology that informates” (Zuboff, 1988). Apparently, our understandings of how work is accomplished in the digitally informated workplace have simply not keep pace with the increased technical complexities. Therefore, as workplace educators we must renew our understandings and adopt more sufficient adult education philosophies and training methodologies if we are to be effective within this maelstrom of technological change.
Philosophies of adult education
If the promise of sophisticated controls systems is to be realized, the operators of those systems must be more highly trained than the “control board” operators of a generation ago. As Zuboff contends: “The cognitive demands of intellective skill suggests that computerization should be accomplished by a serious educational effort, if the informating capacity of the technology is to be exploited.” This is generally unquestioned. However, traditional education philosophies and typical training methods are inadequate for the task.
The first step toward creating more innovative training methods is to examine our own perceptions and assumptions regarding teaching and learning. Our points of view in this regard are heavily influenced by our own formal educational experiences and are deeply imbedded in our personal “frames of reference” (Mezirow, 2000). These fundamental perspectives cannot be easily changed. However, other points of view can be tried-on for size. Most perspectives concerning adult learning can be located within one of the several distinct adult education philosophies. However, the philosophical and theoretical foundations of workplace training are most commonly behaviorist or humanist.
Modern behaviorism was popularized in the 1950’s by its most influential advocate, B.F. Skinner (Elias & Merriam, 1995). It was largely an invention of John B. Watson in the early part of the last century. “As an extreme environmentalist, Watson maintained that he could take any healthy infant and through environmental conditioning produce anything from a doctor to a beggar” (Elias & Merriam, 1995). Of course, most people remember the famous Russian behavioral psychologist Ivan Pavlov and his infamous salivating dog.
Behaviorist training methodologies typically begin by defining behavioral objectives. “Behavioral objectives contain three components: The relevant conditions or stimuli under which a student is expected to perform. The behavior a student is to perform including a general reference to the product of the student’s behavior [and] a description of the criteria by which the behavior will be judged acceptable, unacceptable, successful, or unsuccessful” (Elias & Merriam, 1995). Training is then focused upon presenting and reinforcing desired actions and behaviors. The preeminence of behaviorist methodologies, at least in industrial and military training programs, stems from the generally accepted notion that behavior-based training produces desired outcomes: workers or warriors who perform precisely as trained.
Early proponents of humanism included prominent theologians and educators who were reacting to what they viewed as the dehumanizing elements of behaviorism. Malcolm Knowles is probably the most well known humanist adult educator. Knowles popularized the notion of adult education as “andragogy” as opposed to “pedagogy,” the theoretical constructs and methods employed in the education of children (Knowles, 1980). Although recent research and scholarship make less of a distinction between the learning of adults and children, the central tenants of humanistic adult education philosophy remain consistent. “The student is the center of the process, the teacher is a facilitator, and learning is by discovery…. Humanistic adult education is student centered. In this orientation the teacher does not necessarily know best” (Elias & Merriam, 1995). Humanistic teaching philosophy and methodologies typically find their expression in business and industry in the form of professional, managerial, and executive development programs.
It makes little sense to utilize humanistic methods for management and professional development, and then conveniently relegate worker training to a behavioral domain. Simply teaching to behavioral objectives, no matter how comprehensively defined, will not prepare console operators for the dizzying complexities they face day to day. After conducting a five-year study of the changes wrought by new computer controls systems in process industries, Harvard scholar Shoshanna Zuboff concluded that “The jobs at the data interface become increasingly similar to one another as the informating process evolves. In the advanced stages these become ‘metajobs.’” A constructivist approach of applying varied and even divergent philosophies of adult education may be the most appropriate foundation upon which to construct training for these metajobs. We must understand that “Learners are not just passive recipients, nor are they simple recorders of information....This move away from passive views of learning toward more cognitive and constructivist perspectives emphasizes what learners know (knowledge) and how they think (cognitive processes) about what they know as they actively engage in meaningful learning” (Anderson & Krathwohl, 2001).
Custom high-fidelity simulators
Since the costs and expectations for custom high-fidelity simulation are comparatively high, training designers should commit to a correspondingly diligent effort to design and maintain their simulation training programs. The first consideration is generally the quality of the simulations themselves. Those responsible for the simulation effort must be willing to do what is technically required to make custom simulations work very closely to the way in which the emulated process works. Consider the potential for console operators to make wrong process moves in the “real world” if they have learned on a simulated system that displays different results from process changes or reaction. In short, when creating custom one-to-one simulators, build them right or don’t build them at all.
An important consideration in training design for custom simulation is that console operators do not conduct their business in isolation from others. Decisions about important control moves can involve considerable human interaction, especially during off-normal conditions such as unit start-up. Collaboration between process supervisors, other console operators, field operators, engineering staff, and technicians is dynamic and crucial. Therefore, training activities that are designed to place console operators in real-world situations should include as many human components as practical. Console operators trained within a realistic human dynamic are more likely to engage with others while they make important decisions, and before they initiate critical process changes. Conversely, if a console operator is trained in isolation he or she may approach a real process controls job as a solo activity. This can cause considerable difficulty when teamwork is required. Training activities should simulate as much as possible collaborative decision-making and human interactions. Use scenario-driven role-play as much as practical.
Scenario-driven interactions between key individuals, others, and process simulations are more powerful than is generally understood. They have the potential to motivate real positive change by surfacing real operational issues. Too often process training using even the most sophisticated simulators denigrate into mechanistic drill, drill, drill patterns that only scratch the surface of learner needs. Simulated scenarios should challenge the learner to think about the underlying reasons for their decisions, to consider the ramifications of control moves and motivate them to drill down into deeper understandings of the processes that they control.
Generic process simulators
Generic simulators that emulate typical petroleum refining, chemical, pulp and paper, and a variety of other industry processes can be purchased relatively inexpensively and installed easily. Learners can generally begin their training activities with a generic simulator with little more than a few minutes of instruction. These small but powerful tools can easily be utilized to teach process theory, chemistry of the process, typical process reactions, instrumentation basics, and console display interactions. The cognitive foundation-building that can be accomplished with the use of generic simulation should not be underestimated. These dynamic machines can be considerably more powerful teaching tools than typical CBT’s or “eLearning.” Generic simulations can engage learners, ignite a dynamic learning environment and promote active interaction among participants. In sort, they are practical, powerful and if used creatively, lots of fun.
Generic process unit simulations are typically loaded on networked or free-standing PCs and operated with a mouse and common QWERTY- style keyboards. However, most generic simulation programs can be displayed on proprietary controls systems consoles by linking them together with translator devices, or “black-boxes.” This adds some hardware and licensing cost, but allows for teaching site-specific equipment architecture, proprietary control system interactions, and process control displays. Further, in such a set-up the generic simulations can be shown side by side with the real process unit control displays in a failsafe view-only mode. In this way learners can toggle back and forth between the generic simulations and real-time data, live schematic displays and the distinct configuration screens found within their own process units. This can provide the learner with a significant cognitive-link between the simulated unit and the real thing.
The design of activities for use in generic simulation training can generally be more freewheeling and loose than with custom simulations. Games, role-plays, completion, exploration, and free experimentation can be brought into the training package. This does not mean that the trainer or training designer does not need to be rigorous and to utilize well thought out activities. It simply means that process-basics and console-fundamentals training are much more broadly based and are generally attended by larger numbers of learners with more numerous workstations available. This opens up a larger array of instructional design possibilities. Concepts of teamwork and collaborative decision making can also be taught using generic simulations. Those are not just the purview of those using custom simulations.
Learning styles, preferences and capabilities
The broad categories of visual, auditory, and kinesthetic learning styles are widely recognized. Recommendations to consider different domains of cognitive, affective, and psychomotor domains of knowledge are commonplace. Some learners are self-directing and will eagerly come into a simulation lab and actively explore (Brookfield, 1986). Others prefer pre-set drills and instructor-led activities. Self-paced programmed instruction is milk and honey to some, dry bones to others. Although it may appear that designing training activities that accommodate different learning styles and preferences is unnecessarily complicated and impractical, in practice this is not the case. Training designers can generally accommodate most of the learning styles and preferences of participants by simply making available to them a reasonable variety of training activities and learning resources. One-size-fits-all dogmatic approaches to training design are inadequate and simplistic. Such a myopic view will almost certainly limit the effectiveness of training. In short, mix it up.
Simulators have become more sophisticated, affordable, and commonplace. However, methodologies to use simulators effectively to teach process plant console operations have not evolved sufficiently to exploit their power. For this reason industrial plant landscapes are replete with the carcasses of dust-covered simulators, abandoned and useless. If we as 21st century workplace educators are to be successful in our charge to train and equip the post-modern workforce, we must first reflect upon our own perspectives and philosophies regarding teaching and learning. If necessary, we can then choose more adequate foundations upon which to construct our training. We must understand that simulators are not training machines that can simply be plugged-in and expected to teach people. People teach people, by using simulators.
We should research, consider, find, or invent a variety of ways to utilize these sophisticated simulators to teach process control, and then put them to work. Finally, if we can be as innovative in teaching digital control systems as those who invented them in the first place, we will see the promise of dynamic process simulators fulfilled.
Dr. Richard C. Ortloff has been a training practitioner in the petroleum and petrochemicals industry for more than 15 years, and is currently a senior training advisor for ExxonMobil in Beaumont, TX. He has a Doctor of Education degree from Columbia University in New York. In 2001, he was certified by Honeywell Automation College to teach Console Operations. Dr. Ortloff was also a full-time process technology instructor for two years at a Pacific Northwest Technical College.
Anderson, L. W., & Krathwohl, D. R. (Eds.). (2001). A taxonomy for learning, teaching and assessing: A revision of Bloom's taxonomy of educational objectives. New York: Longman.
Brookfield, S. (1986). Understanding and facilitating adult learning (Vol. 375). San Francisco: Jossey-Bass.
Elias, J. L., & Merriam, S. B. (1995). Philosophical foundations of adult education (2 ed.). Malabar, FL: Krieger Publishing Co.
Knowles, M. S. (1980). The modern practice of adult education: From pedagogy to andragogy. Engelwood Cliffs, NJ: Cambridge Adult Education.
Mezirow, J. (2000). Learning to think like an adult. In J. Meziorw (Ed.), Learning as transformation (pp. 3-33). San Franscisco: Jossey-Bass.
Zuboff, S. (1988). In the age of the smart machine: The future of work and power. New York: Basic Books.
For more information, visit:
For more reading on simulators, visit controleng.com/archive:
- Events & Awards
- Magazine Archives
- Digital Reports
- Global SI Database
- Oil & Gas Engineering
- Survey Prize Winners