Cognitive technology development and end-user involvement in the Norwegian petroleum industry – Human factors missing or not?

• There are hardly any involvement of end users in early phases such as designing algorithms and training data of cognitive technology development in the Norwegian petroleum industry. • The reason for this lack of involvement include limited access to end user and that there are no regulations on the involvement during early phases. • A lack of human factors methods for early technological development contribute to little and low quality end user involvement in the early phases of complex technology development. • The findings are very important for development of regulations and practical advice on a how to achieve a broader inclusion of end users in all phases of complex technology development for increased safety during operations. The petroleum industry is a high-hazard industry depending on reliable technical solutions. The industry tends to use increasingly advanced technologies including machine learning technology with increased difficulties for end users to keep abreast of how these technologies work. Thus, our research question was: How are end users involved in the development and implementation of cognitive technologies in the Norwegian petroleum industry to contribute to safe and reliable technical solutions? We used a qualitative explorative approach, with semistructured interviews with 31 informants from 10 companies. Thematic analysis revealed the categories 'technology focus', 'understanding of end-user involvement versus end users' actual involvement', 'lack of access to end users', 'lack of human factors methods in early phases', and 'lack of official rules and regulations'. Findings show that during the earlier phases of designing algorithms and training data, end users are hardly involved. Regarding later phases with offshore testing, implementation, use, and improvement, end users are much more integrated in the process. [ABSTRACT FROM AUTHOR]