That was one of the questions that puzzled me while I was still a student. For quite some time I had trouble relating what we learnt in our syllabus with the type of activities you would be doing in your day-to-day job.
Granted that in our last year we had a few real case investigations that had been carried out in the lab, and we had a few factory visits to Sidmar, but all of these were cleaned and polished to such an extent that it seemed as if in most instances the processes looked after themselves, and whatever deviations there were was up to the operators in the pulpit to deal with.
And maybe that’s why being a researcher at university was not considered a proper job as seen from the point of view of a person in the steel industry: it was really a cushioned life which possibly exposed you to some technical aspects of the job, but not to the real grit of an engineer’s job.
Because once you’re there doing the real job of an engineer, you learn pretty quickly how sanitised the processes in your syllabus were, and that in the real world you need a team of people, not just to develop the process for future requirements, but also to make sure that deviations from the desired process flow are properly understood and then dealt with. Otherwise you just put a plaster over the problem but you haven’t really addressed the underlying root cause.
In the end, you also learn that being an engineer is not just about being a technical boffin, but also dealing with people: especially as a junior engineer you have to try and convince people whose first job is to produce the tonnage that they should allow you some time to perform whatever experiment you want to perform on the live mill. So you quickly learn how to influence people without having any authority over them. In fact I became so good at this approach that I had trouble on the occasion that I had to lead a team of people and try and make them do what I wanted them to do without being overbearing or too wimpish.
So, to answer the original question “What does an engineer do?”, the answer has to be: anything that is needed to make the process in the here-and-now behave as closely to what the books say it should do. Chemical reactions may look after themselves, but someone has to make sure that the conditions are correct for those chemical reactions to follow the desire path.