Hi-I am doing a PhD in an engineering-related subject-my supervisor expects me to build a piece of software by the end of 3 years, and tells me that I will get a PhD based on mainly on the software,with some experiments evaluating the software on the side.As far as I was aware,people are awarded a PhD based on research, and not so much on a final product.Should I be examining another area related to my paid project for purposes of the thesis?My main worry is that the software may not solve all the problems that it is being developed to address.In that case, would that lead to a fail in the PhD?Could anyone provide me with some advice?Thanks.

Hi, I am not in an engineering field but there are common themes to all PhD research and from what you’ve said, I don’t think you need to worry. The aim of your project is to design some software to achieve a specific task, and then to beta-test it? Your research element will be in working out the best way to create the software and also, the best way to validate it. So I think you will be awarded your PhD based on your journey from initial idea of designing the software, through to building and evaluating it. As with any project you’ll defend the way that you achieved this and reflect on how it might be improved. If your software does not function in the way it should (and so, does not fulfill the original aims of the project) then this would not mean PhD failure as long as you can defend the choices you made along the route from design to final product. I hope that offers you a bit of reassurance but to put your mind at rest, I think you should talk to your supervisor about your concerns. Best of luck with your work.

Thanks Ann.I really appreciate your advice and reassurance - its cheered me up :) I have spoken with my supervisor who tells me a software based approach is what he wants,and the funding body is expecting a product by the end of the PhD. Other researchers have told me that this approach may be dangerous if the methods taken to develop the software are not valid, but I guess I have to find a balance between the software development and theories behind development/evaluation.