Confidentiality of thesis data: PhD with industrial partner

Question

In the field of Mechanical Engineering, when doing a PhD in a University but in partnership with a company, at some universities, an agreement is signed between the company and the university (or Laboratory) in which the company remains the owner of all files provided (e.g. geometries, meshes) as well as the results produced, but allows for publication of most of the results. The line drawn between what its supposed to be published and what remains strictly confidential is hard to draw and often depends on the company as well as the topic.

Usually, in order to preserve this confidentiality, researchers scale the figures to unknown values and do not provide geometries or mesh/material properties for example. I was wondering if this common practice in scientific publication can be applied when writing a PhD Thesis. Basically, are all results presented in a thesis supposed to be public and more importantly, is an independent/external researcher supposed to be able to reproduce all of the results ?

Answer 1

The standard for a Ph.D. thesis is simple and universal: it must make a contribution to the sum of human knowledge.

As such, all results in a thesis should be public and presented in enough details to allow another researcher with sufficient background to reproduce them.

It's fine for some of the work that you are doing to not be public, just so long as enough of the material is public to make up a Ph.D. thesis. For example, if you have figures scaled to unknown values or missing geometry information, then you need to ask: is there significant scientific value even without this information? For example, if the proprietary figure is an example of how a method was applied, that's probably OK as long as there are other forms of verification elsewhere. If the proprietary figure is the verification, however, and it's missing critical information, then that is not OK.

Answer 2

The answer to this question is going to vary from university to university. At our university thesis can be placed on a restricted access list, and examiners made to sign confidentiality agreements. Agreements I have signed to specify a time limit on how long the thesis must be kept confidential, but do say that the commercial partner will indicate when they no longer need access to the thesis to be restricted.

Once the thesis is published, it is entirely possible that materials that cannot be disclosed may have been used in the thesis. For example, one would not include code for proprietary routines, but would rather say "the XXX function from the library provide by Blogs Ltd". I guess it would then be up to the examiner to judge if enough information had been provided that they could judge the value of the work. It may be that, since the examiner has signed a confidentiality agreement they can be made privileged to information that other readers cannot. In my experience these issues are sorted out on a case by case discussion between the IP holder, the university and the examiner.