I earned a degree in engineering physics, a now-defunct program at Kansas University created for nuclear submarine captains. Doing double duty, I took the required courses for a physics degree, plus 32 hours of engineering electives. With job offers from aerospace, chemical, and communication companies, I chose the chemical company because the campus was impressive, and I would have my own cubicle, but I had no idea what an instrument engineer did. I had never even heard of the profession until my interview.
I didn’t expect to use my courses in quantum mechanics, statistical physics and astrophysics, although I would later see The Uncertainty Principle, Spin, and Relativity in executive decisions. I was unprepared for the fact that nothing I learned in getting my degree was used on the job except for a few concepts from courses in heat transfer and fluid mechanics and an electronics lab for test instrumentation.
Concept: Success as an automation professional requires broad technical knowledge. You are a generalist dependent upon other people’s expertise that is gained on the job and largely undocumented. Communication and interpersonal skills are more important than ever. Good conversation can be mutually beneficial in terms of broadening horizons.
Details: Stop by the office of process, mechanical, E&I, and configuration engineers in your plant—or corporate offices—with an intelligent question. First, do your homework by talking with associates, searching the Internet, and checking out handbook sections on the subject. Armed with knowledge, call the corporate office of your supplier and ask for technical support. Follow up with anyone who gives you advice. Engineers love to solve problems—take advantage of this. Get to know the technicians in the instrument shop. Ask questions about what works and what doesn’t. When in a plant, spend each morning in a control room to see what the operators are doing, and become aware of any problems they are having with loops. Review trends for oscillations and trajectories. Patterns are often more apparent in the PID output since the PID is transferring variability from the process variable to the output. Group all the process and utility flows on the same trend with a time scale of about 20 deadtimes of the slowest loop to see where changes start and how they propagate. Check these trends for different shifts to see if there are operational differences. Ask the operators what is occurring during the worst trends.
Watch-Outs: Experts can be wrong. Never discount the possibility that some aspect outside an expert’s knowledge may be at play. Provide polite feedback, but do not try to change the expert’s mind, and do not hesitate to seek advice from the same expert for the next problem. Do not expect expert systems to capture this expertise. Expert systems were the rage in the late 1980s. Nearly all expert systems in the process industry went out of service often as soon as the expert left the control room. Human expertise is often more fuzzy, disorganized, intuitive, and heuristic than analytical. Incessant chatter and a few false alarms ended up getting these systems turned off sooner than later.
Exceptions: Do not ask maintenance questions during a turnaround or ask operations questions during a shutdown, startup, or a flurry of alarms.
Insight: Engineers, technicians, and operators will tell you what they know if your questions are intelligent and non-confrontational.
Rule of Thumb: Start a conversation with everyone responsible for the successful implementation, maintenance, and operation of your designs with a friendly, open, humble, and patient approach.