Governing AI Projects in the Manufacturing Industry : From successful pilots to high-value tools
Lehtinen, Juhana (2026)
2026
Tuotantotalouden DI-ohjelma - Master's Programme in Industrial Engineering and Management
Johtamisen ja talouden tiedekunta - Faculty of Management and Business
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2026-02-04
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202602042273
https://urn.fi/URN:NBN:fi:tuni-202602042273
Tiivistelmä
Artificial Intelligence (AI) has emerged as a significant source of value for industrial manufacturing companies by driving efficiency and innovation. However, many organizations struggle to scale tools to their full potential, a phenomenon known as the pilot paradox. Even if an organization has technical capabilities, the underlying structures required to support AI remain underdeveloped. The objective of this thesis was to create a framework that can be used to effectively understand and manage AI systems in the context of a large international industrial manufacturing company.
This study began with a literature review to build a theoretical foundation for the research. Three distinct dimensions were identified: strategic alignment, AI governance, and lifecycle management emerged as the three main themes for future research. Second, a single-case study of an ongoing AI project was conducted to understand practical implementation strategies and real-life phenomena. Third, semi-structured interviews were conducted with IT and business professionals at the case company. The interview data was then processed using the Gioia methodology to understand the connections and mechanisms associated with pilot paradox.
The results show that the pilot paradox is caused by strategic ambiguity, evolving competence, and fragmented governance. Even if pilots meet the standards from a technical perspective, they often fail to scale because of undefined ownership and a lack of formalized processes for the maintenance of tools. The study also highlighted the urgency of the technical Machine Learning Operations (MLOps) engineer role, responsible for the technical side of AI systems.
This thesis culminates in a management framework that includes governance checkpoints for strategic alignment before a project is initiated. Central to the management framework is a RACI matrix, which explicitly divides IT and business responsibilities into different parts throughout the project lifecycle. This research contributes to the field of AI management by synthesizing the existing literature on the topic and creating new knowledge via the case study and interview round.
The research shows that effectively scaling AI tools to their full potential requires rigorously defined processes, disciplined governance and technical capabilities. The final framework acts as a broad guideline for IT and business professionals to handle the challenges of scaling AI tools to their full potential.
This study began with a literature review to build a theoretical foundation for the research. Three distinct dimensions were identified: strategic alignment, AI governance, and lifecycle management emerged as the three main themes for future research. Second, a single-case study of an ongoing AI project was conducted to understand practical implementation strategies and real-life phenomena. Third, semi-structured interviews were conducted with IT and business professionals at the case company. The interview data was then processed using the Gioia methodology to understand the connections and mechanisms associated with pilot paradox.
The results show that the pilot paradox is caused by strategic ambiguity, evolving competence, and fragmented governance. Even if pilots meet the standards from a technical perspective, they often fail to scale because of undefined ownership and a lack of formalized processes for the maintenance of tools. The study also highlighted the urgency of the technical Machine Learning Operations (MLOps) engineer role, responsible for the technical side of AI systems.
This thesis culminates in a management framework that includes governance checkpoints for strategic alignment before a project is initiated. Central to the management framework is a RACI matrix, which explicitly divides IT and business responsibilities into different parts throughout the project lifecycle. This research contributes to the field of AI management by synthesizing the existing literature on the topic and creating new knowledge via the case study and interview round.
The research shows that effectively scaling AI tools to their full potential requires rigorously defined processes, disciplined governance and technical capabilities. The final framework acts as a broad guideline for IT and business professionals to handle the challenges of scaling AI tools to their full potential.