I would say that engineer teacher should draw upon the Bloom's Taxonomy of Learning. Assessments are always done against the learning objectives of a concept/topic/course/program. While defining the objectives, various levels for desired learning should be kept in mind and assessment instruments (exam papers) should be designed in line with those objective. If a learning objective requires that the learners must be able to understand the functioning of a particular machine/equipment then the learners must be tested in way that ensures the assessment of the level of this ability which can only be done through some practical problem rather than an oral or written answer. Understanding of Bloom's Taxonomy and it's variants (Google these key words for details) can be a good starting point.

I would make an exemplary glossary for the course. I would then analyse the ontology of the glossary.

I would then produce a concept map for the whole course.

I would analyse the highest (important) concepts or the longest (big) threads in the concept map and examine these.

For example the important concepts "Research" and "Data" are related through the verb "needs". The student can then be asked to fill in the missing word: "Research ____ Data". (We are in fact doing all this for my course.)

It is very interesting approach, I will try to embed it in my practice. Thank you!

To endure that students are not rote-learning, it is useful to get them to demonstrate that they can apply a Concept-Relation-Concept.

I would say that engineer teacher should draw upon the Bloom's Taxonomy of Learning. Assessments are always done against the learning objectives of a concept/topic/course/program. While defining the objectives, various levels for desired learning should be kept in mind and assessment instruments (exam papers) should be designed in line with those objective. If a learning objective requires that the learners must be able to understand the functioning of a particular machine/equipment then the learners must be tested in way that ensures the assessment of the level of this ability which can only be done through some practical problem rather than an oral or written answer. Understanding of Bloom's Taxonomy and it's variants (Google these key words for details) can be a good starting point.

If it is not examined, students will not learn it!

I have not read yet Bloom's Taxonomy - I will do it soon, thanks for pointing me out, Rauf. Ian, I absolutely agree!

I would also suggest application of the concept as part of the assessment. Coming from teaching in an engineering University, I always focus on concept application as a metric of concept understanding. I always say that the student can have a some understanding of what are the concepts of an operating system in a computer, but he has not grasped the whole idea behind the concept until he has learned to program it. The same goes with civil, mechanical, chemical and electrical engineering in general.

I would suggest a two pronged approach based on Ian's and Rauf's suggestion coupled with practical assessment of concept application