Ghanem, AbrahamHerrmann, Andrea2023-11-302023-11-3020230720-8928https://dl.gi.de/handle/20.500.12116/43240Machine Learning (ML) based industrial applications deployed in high variance dynamic environments present a new set of challenges. The performance of such systems is directly linked to the nature of the data it has been subjected to. Using the computer vision-based ML applications in the logistics industry as a case study, due to their high variance environment and strict requirements, we try to address the issue of understanding the data requirements for the successful development and deployment of such applications. We propose a systematic approach to address high variance scenarios with limited relevant data availability, covering both real data collection and synthetic data generation, highlighting their requirements and effective utilization methods.enMachine Learninglogistics industrycase studydata requirementsText/Conference Paper