Open Access

Timo Schroth, Vera HummelORCiD, Jan SchuhmacherORCiD, Konrad von LeipzigORCiD

• Small and medium-sized enterprises face challenges in the performance-oriented improvement and integration of digital solutions in pick and stow operations, due to a high degree of manual processes. Traditional methods for improving these processes based on historical data are becoming increasingly ineffective due to short product life cycles and small order quantities. The use of AI offers great potential for data-driven analysis and optimization of pick and stow operations. Currently used AI algorithms focus on optimizing either pick or stow operations but not both in combination, missing the opportunity for holistic improvement of warehouse operations by reducing walking distances and process times. This paper targets these limitations by investigating AI algorithms for near-real-time AI-based clustering and combination of pick and stow operations conjointly. The research addressed in this paper aims to review AI algorithms for the clustering and combination of pick and stow operations to set the basis for the enhancement of these algorithms by incorporating close-to-real-time data analytics to improve logistics performance. The findings of these investigations will be transferred directly into an undergraduate learning module for first-year students to give them a basic understanding of the possibilities of using AI for clustering and combination of pick and stow operations, hence preparing these students for their first industry internship. In addition to a theoretical introduction, the concept includes practical holistic scenarios in the Werk150, the learning factory of the ESB Business School on the campus of Reutlingen University.