Populating the Digital Twin of a Retail Store Using an RFID Autonomous Mobile Robot

In this paper we describe a real case deployment over several months of an autonomous mobile robot equipped with RFID sensors, able to populate a digital twin with the inventory and location of hundreds of thousands of products in a very large retail store in the USA, comprised of several floors and thousands of square meters in area. We will also cover all the practical details involved in deploying this solution in the store for several months, including how the robot adapts to changes in the layout of the store, how the performance of the nightly mission can be monitored and evaluated remotely, and how the data is used to populate a digital twin of the retail store. Finally, we will present the principles used in the algorithms used to determine the inventory in real time, and to estimate the location of each product using a combination of the robot navigation data and the RFID readings.