Russian scientists have proposed a new method for analyzing customer behavior using a robot that carries out autonomous inventory using radio frequency tags (RFID). The research results were published in the materials of the ICARCV conference, the press service of the Skolkovo Institute of Science and Technology writes briefly about this.

The robot, named Michelle, monitors shoppers, determines the most attractive product locations for them, and predicts demand. It is equipped with an array of RFID antennas, thanks to which it can read data from radio frequency tags with which the goods are equipped, and transmit the result of the inventory to the warehouse management system.

“Existing solutions do not apply to real-life retail situations, which could result in unexpected loss of sales. The solution we propose provides comprehensive information on the distribution of demand in the sales area using a mobile robot for autonomous inventory of stores in which goods are marked with RFID tags. Our research differs in that we operate with a large amount of initial data collected in a real environment,” explained one of the developers, Alexander Petrovsky.

Using the method developed by scientists, you can build models that reflect the dependence of demand on the location of goods, and determine the areas from where goods are least likely to be purchased, as well as find lost and moved items.

The robot can significantly reduce the number of errors in reading RFID tags caused by human factors, as well as speed up and reduce the cost of the inventory process. The developed approach is applicable to any sales area where goods are marked with RFID tags.

“The results obtained are very important for retailers in terms of finding better product placement schemes in order to maximize sales profits and predict seasonal demand dynamics,” said Dmitry Teteryukov.