CC BY 3.0 UnportedSchuh, GüntherGüntherSchuhCassel, LeonardLeonardCasselMichels, MaximilianMaximilianMichels2025-07-282025-07-282025-07-14https://publica.fraunhofer.de/handle/publica/489991https://doi.org/10.24406/publica-495510.15488/1902410.24406/publica-4955The manufacturing industry is undergoing a fundamental transformation, driven by the ongoing digitalization of all stages and processes across value chains. The extensive collection and subsequent automated analysis of data with the help of artificial intelligence (AI) not only enables a higher degree of automation of manufacturing processes, but also a significant increase in their efficiency. AI pilot applications are increasingly being brought into industrial settings, demonstrating the potential benefits of adopting AI as a technology for production systems. However, pilots are primarily being trialed either by major companies with access to vast resources or by research institutions. In contrast, small and medium-sized companies are faced with the challenge of identifying the most beneficial uses of AI applications for their individual production systems while facing limited resources for the actual implementation. An objective assessment of the cost-benefit ratio is required to select and implement the most promising use cases. In addition, interactions between decision-relevant parameters must be considered in the selection process, which are often only recognized in the course of implementation. This paper aims to identify and evaluate value contributions of current AI applications in production. A literature-based assessment using the PRISMA method encompasses discriminative AI use cases in the manufacturing industry and highlights distinct types of value contribution with a focus on the main dimensions time, costs, quality and flexibility.enManufacturing industryIndustrial productionArtificial intelligenceValue contributionsMachine learning600 Technik, Medizin, angewandte Wissenschaften::620 Ingenieurwissenschaftenjournal article