Options
2027
Conference Paper
Title
Developing and Configuring Artificial-Intelligence Systems in Manufacturing: Adoption Insights from a Practitioner Survey
Abstract
Artificial-Intelligence (AI) systems are increasingly regarded as key enablers of smart manufacturing, offering benefits such as reduced downtime, improved efficiency, or enhanced quality. Yet, the adoption of scientific approaches in practice remains limited due to persistent challenges in appropriate transfer, implementation, deployment, and configuration. In this article, we report on a survey with 26 AI-system practitioners from the manufacturing domain, focusing on their opinions and experience regarding three pillars: development, configuration, and perceived value. Our results show that development is primarily hindered by data quality and integration barriers, while configuration problems, such as version conflicts or unintended feature interactions, occur frequently and lead to business-critical impacts, including downtime and vulnerabilities. Nevertheless, practitioners perceive clear value in AI systems, especially in predictive maintenance, although broader applications, such as real-time optimization or advanced quality control, remain underexploited. Overall, the study provides valuable insights into how AI systems are currently adopted in manufacturing practice. The findings highlight the need for configurable, reusable, and scalable solutions, as well as systematic data management and configuration practices to facilitate sustainable adoption in manufacturing settings.
Author(s)