Customer behavior analytics using an autonomous robotics-based system

Abstract

This paper suggests a novel method for customer behavior analytics and demand distribution based on Radio Frequency Identification (RFID) stocktaking. Existing solutions lack applicability to real-life situations in retailing, which may result in unobservable loss of sales. The proposed solution provides new parameters of demand distribution to the retailer using a mobile robot for autonomous stocktaking of RFID-equipped shopping rooms. Built models depict location-related demand dependencies, the most and the least purchasable areas in a store, and precise localization of lost and moved items. Our research differs from the related works by the sheer size of the underlying data set collected in a real-world environment for more than ten months.