In my prior post, I proposed that each engineering position requires a different level of abstraction. To a research engineer, almost everything is model-based, while the production engineer may be primarily focused on issues that are object-based. Although freshman and sophomore engineering students receive guidance as to the sub-discipline they should enter (electrical, mechanical, chemical, etc.), I’ve never seen any discussion about specializing in a particular level of abstraction. So I want to illustrate how skill sets vary according to one’s position along the engineering spectrum. The following abbreviations are used below: high abstraction (HA), moderate abstraction (MA), and low abstraction (LA).
Solution focus
HA: Primary focus is finding an optimal solution within a tightly controlled problem domain. Journal referees don’t want to read about yet another mediocre solution; they want to see mathematical, statistical, or experimental evidence that the proposed solution is in some manner better than previously discovered approaches. Only a single solution can be considered best.
MA: Central effort is placed in discovering a bounded solution. For instance, a bridge doesn’t have to be optimal in every respect, but it had better withstand the specified traffic loads. A bridge with too much strength is of far less concern than one with too little carrying capability. Any solution that meets the project constraints is potentially useable.
LA: Making sure that each component/batch/output is operating correctly often requires a rapid solution. If a manufacturing process is going out of tolerance, the first concern is getting product back within tolerance. Causes of the deviation can be examined later, or passed on for further study, but the key focus is on quickly finding a solution that works. For outputs of sufficient financial worth, almost any workable solution will be considered acceptable, at least on a temporary basis.
Solution domain
HA: Solutions are developed in the symbolic domain, where analytic tools of mathematics are most effective.
MA: Problems are solved in the spatial or schematic domains, where computer-aided-design (CAD) tools allow the consideration of multiple solution possibilities.
LA: Troubleshooting success is highly dependent upon prior exposure to similar problems, and thus the requisite skills are experiential in nature.
Social influence
HA: Symbolic solutions stand on their own, and require minimal social interaction to be presented and accepted.
MA: Gathering problem specifications, managing organizational expectations, and presenting solution proposals requires a moderate level of social interaction.
LA: Talents in motivating and managing others are quite valuable in bringing the right technical skills to bear on a problem, and in coordinating troubleshooting activities, especially in a high-pressure manufacturing environment.
Temporal effects
HA: Symbolic solutions do not care when a system is set into motion, as nature’s laws are assumed not to vary with the passage of time. Thus, problems of high abstraction accommodate everlasting solutions.
MA: Schematic solutions can remain valid over long periods of time. However, as types of components or methodologies change with time, moderate abstraction problems may need to be updated and improved.
LA: Corrective solutions may be specific to particular outputs, or a specific set of events acting on an output. Thus, actions associated with low abstraction problems are highly time dependent.
Summary
Individual engineers may have to move up and down the engineering spectrum over the course of a career, or a year, or even a single day. This post has attempted to point out that the skills needed to be a successful engineer necessarily vary with the abstraction level being utilized. In my next post, I’ll discuss why most engineering students are only exposed to high abstraction skills during their time in college.