In the engineering sciences, a standard is something that is arrived at the end of a prolonged process. Standards can, for example, define procedures, workflows, formats, boundary conditions, characteristics or reference values, but they must always be critically questioned and if necessary further refined. It is also possible that several standards might apply to the same problem or question.

Establishing standards takes place within subject-specific communities, but it is not usually the primary goal of research. Instead, it tends to be a subsequent step towards general, broader application as part of the transfer of research results.

In the context of research in the field of engineering science, it makes more sense to regard a standard as a point of reference – a reliable benchmark that enables classification of research results in relation to the state of the art. Reliable comparisons might include experimental approaches and algorithms recognised within the subject-specific community or data from benchmarks.

References and the standards derived from them can be defined at different levels: as a minimum requirement for a working methodology recognised in the field, as a reference model that has already been elaborated or, at the highest level, as a formalised standard adopted by organisations such as ISO or DIN.