When a subscribed user enters a shopping area (e.g., a mall), he is identified and his location is continuously tracked. Based on his profile and movement, the system generates smart retail recommendations delivered through the app on his device, also using AR services. The Smart Retail application processes data at the edge by collecting information from distributed sensors (e.g., Bluetooth beacons) and from user devices. The device reports real‑time position, time spent in each store, and possible sensor readings. Data analytics and ML techniques, optionally supported by cloud services, determine the contextual audio/video information to send to the appropriate edge node for AR rendering.

All collected data also feeds a database storing previous interactions, such as past purchases.

The application includes a central component reachable via a fixed URL, contacted by visitors’ devices, which redirects them to the most suitable edge instance. The solution is deployed across multiple network domains and consists of several software components, some of which can be dynamically moved to the edge depending on user density in the mall.