Health Assessment of Inventory Records at Assembly Line through Data Analytics

Abstract

Big-data and data analytics are today seen by many industry experts and researchers as key in creating competitive advantage. Insights once hidden in vast amount of data are through digital tools now becoming available. Traditional manufacturers are today realising the necessity of data-driven decision-making, there is however a lack of knowledge of viable use cases. Inventory record inaccuracy (IRI), the mismatch between the recorded inventory and the physical inventory, is a common issue for manufacturers. The literature have concluded that there are multiple causes to the problem, but there is a lack of research on the impact of manufacturers. The characteristics of the problem of IRI makes it highly suitable for utilising the tools of data analytics in order to better understand the issue. However, to our knowledge, there is a gap in the literature on how to utilise data analytics to tackle such a problem. Data-driven decision-making heavily relies on the available data for reliable and valuable output. Organisations are therefore further challenged to find ways of assessing their data’s quality to ensure the correct output. This thesis aims to investigate the impact of IRI at a manufacturing company, and analyse the extent of the problem through data analytics. Additionally, the thesis aims to propose how data quality in organisations should be assessed to facilitate future data analytics projects. A mix of quantitative and qualitative methods were used to generate the results. Interviews, observations, and data from existing databases of the company were used as the primary sources for data collection. The findings indicate that the impact of IRI is severe and affect multiple functions of the investigated company. Further on, IRI is identified to have direct and indirect consequences, where the latter are found to have the greatest impact on the company. Data analytics is demonstrated to facilitate quantifying the extent of IRI, furthermore, the comparative deviation demonstrates a strong foundation for further categorisation of inventory record deviations. To facilitate future data analytics projects, organisations should centre their assessment of data quality around the data consumer, and through frequent reviews, aim to constantly strive for better alignment between their data and the data consumers’ needs.