I mostly know about shipping and petrochemicals, which are at least somewhat specialized, so I could be wrong about other areas. I personally wish everything were more abstracted, because in applied AI work (my area) the lack of uniform abstractions where we can just drop in off the shelf stuff to solve a company's problem is a big issue!

It's not so much that the techniques are completely different, as that people who work in the area use a set of abstractions built for that area, and are expected to know those particular abstractions, along with the terminology, typical practices, and constraints on why things are done a certain way. Maersk logistics engineers have a set of approaches built up over the years relating to the economic, legal, and technical context of shipping, for example; there's a whole pile of domain knowledge there, in addition to the general knowledge.

In petrochemicals some chemical-engineering experience is often preferred as well, since optimizing a refinery complex doesn't always break down cleanly between logistics and process-engineering specialties, so you often want people with backgrounds in both.