Well you could apply an awesome AI method to an important problem and report your results. Then you would have first hand knowledge of what the field is capable of.. That would be my suggestion for a dissertation. You could also discuss this with your advisor who probably has some interest in the topic. Best wishes, David Booth

You probably will have more success with a small number of in-depth case studies than with a survey, Sebastien Vandendriessche .

With some exceptions, case studies and related qualitative research help you find out what is going on, and possibly why. A survey helps you estimate how much it's going on.

A survey assumes that you and your respondents all understand terminology the same way. What you mean by AI may not be the same as the managers at your manufacturing companies mean by AI. For example, a robotic arm that sweeps away cookies that are the wrong shape is a form of AI.

To begin, you will need to review the many forms of AI that may be used in different manufacturing applications. You may then decide to confine your research to certain types of AI and certain types of manufacturer. Then you could make contact with senior managers and interview them about what the decision process was, the extent that efficiencies were sought, the factors that helped or hindered the project's success. While they may not be directly related to AI, there is a lot of research about organisational projects and how they can fail or succeed as a result of organisational factors - communication, buy-in from stake-holders, planning, etc.

After you have a good idea about what AI, what type of manufacturing, success (profitability) factors, and examples of each, then you could consider a survey. Or not. (I'm not saying surveys are bad - I've spent my career dealing with surveys and survey data - I'm saying that good surveys require that you have a sound understanding of what you're trying to find out.)

I guess that there are two problems here:

If you want to study problem 1, then you can start with the review paper provided in the link. It discusses all the major maturity models. They have also discussed some empirical papers so you can get our maturity questionnaire from those and adapt them to your study

On the other hand, you may want to look at the financial performance of companies that have used AI. You can refer to some papers that have looked at a similar approach of 'effect of XYZ' on the financial performance of firms. Some papers derive data about the use of technology from the annual statements, they use some other variables and regress it against the indicators of financial performance.

You may use the technology maturity of AI as one of the antecedents of firm performance. However, I recommend that you focus on either one of these to make things simpler. Each of these two problems deserves a separate paper by itself.

Article The maturity of maturity model research: A systematic mapping study