Purpose – the purpose of this study is to investigate managerial and organizational barriers that are associated with artificial intelligence (AI) and develop a structured process to overcome the organizational barriers throughout different phases of the implementation process. Method – The study applied a qualitative research approach that consisted of multiple case studies from various organizations in traditional industries. Each organization worked with an AI-projects that were based on application of machine learning (ML). The respondents came from various positions from the AI-project and were interviewed. The collected data was analyzed by using a thematic analysis with 17 interviews in total. Findings – The study found four barriers in total from pre-implementation, implementation, and post-implementation phases of AI. These were: lack of use-case definition, low ai-knowledge, missing appropriate data, and end-user misalignment. The study would present key-activities to overcome the AI-barriers categorized that are presented in three phases: defining AI-transformation, anchoring AI-implementation and optimizing AI-usage. Theoretical contribution - Firstly, the study highlighted underlying implementation barriers in traditional industries that were business and managerial related. Secondly, the study contributed with an empirically rooted structured AI-implementation process framework. These findings extend current dialogues in the literature on challenges related to AI and connect them to specific phases in the AI-implementation process. These findings also extend current dialogues in the literature on challenges related to AI and connect them to specific phases in the AI-implementation process. Practical Implications - The practical implication of this study highlighted that there existed a lack of clearly defined strategies for implementing AI-solutions in traditional industries which this study covers by developing a basis to build on. Limitations and Future research - The study investigated a handful of organizations in different industries. Because of time- and resource constraints, increased research scope could provide more insightful perspectives which could be beneficial. In addition, because the study itself was based as a qualitative study, the methodology of the project could be prone to inconsistencies or a lack of coherence. As this approach was based on phrasing, some phrases may not be able to capture the full meaning of what was articulated. For future research proposals, a quantitative research method of this subject could give further breadth to the literature by investigating likely correlations.
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:ltu-91756 |
| Date | January 2022 |
| Creators | Drmac, Filip |
| Publisher | Luleå tekniska universitet, Institutionen för ekonomi, teknik, konst och samhälle |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0025 seconds