What is it about?
The paper addressed the research challenge of detecting misplaced products in a retail store (i.e. products that are placed somewhere in the sales floor area and are not where they ought to be) without planogram information (i.e. without layout plans that show in detail where specific product types shall be placed). The proposed approach is called MiProD (misplaced product detection), and its goal is to detect products misplacements by relying only on potentially noisy RFID sensor readings. To achieve this goal, MiProD systematically compares and improves four different analytical methods.
Featured Image
Why is it important?
We investigate the accuracy of four different methods to detect misplaced products, using simulation and a case study from a European fashion retailer. Our results demonstrate the feasibility of misplaced product detection without planograms, both in the simulated environment and in the industrial case. Also, results from the case study suggest that MiProD can achieve a suitable level of accuracy in everyday fashion and apparel retail operations.
Perspectives
In my opinion, the paper is both timely and relevant for researchers and practitioners. The first ones may benefit from the results of our experiment, which investigated in how far misplaced products can be detected using outlier detection techniques without having a planogram available. These results have implications for research into sensor based locating systems, misplaced product detection, and for their joint application in practice. Our work has also implications for practice. The results clearly demonstrate the potential of improving the analysis of the raw data provided by RTLS. Vendors of such systems might be better advised in fine tuning their analytical software than investing in more powerful hardware. Our results show that accuracy can be achieved without having to rely on planogram information.
Mr. Giovanni Romagnoli
Universita degli Studi di Parma
Read the Original
This page is a summary of: Misplaced product detection using sensor data without planograms, Decision Support Systems, August 2018, Elsevier,
DOI: 10.1016/j.dss.2018.06.006.
You can read the full text:
Contributors
The following have contributed to this page
