RFID self checkout in the FMCG industry: an experimental approach

This study experimentally evaluates the effectiveness of RFID technology for self-checkout applications in the Fast-Moving Consumer Goods (FMCG) sector, aiming to identify the key factors affecting reading accuracy. An experimental campaign was conducted using a fractional Design of Experiments (DoE) approach, considering shopping cart material, product quantity, presence of liquid products, presence of metal-containing products, and use of shielding bags as independent variables. Reading accuracy was analyzed through ANOVA to quantify the influence of physical and operational factors.
The results show that reading accuracy ranges from 73% to 100%. The presence of liquid products emerges as the most significant factor, causing the largest performance degradation. The impact of metal-containing products, although difficult to isolate due to the pervasive presence of metal in carts and shielding bags, is statistically significant and mainly results in an increased performance variability. Product quantity exhibits a negligible effect. The highest accuracy is achieved under favorable configurations, particularly in the absence of liquids. Significant interaction effects are also observed, especially between cart material and the presence of metals.
The study provides a quantitative assessment of RFID self-checkout performance in realistic FMCG environments, highlighting its current technical limitations, and offering practical insights for technology developers and retailers.