An Interactive Cart with Analytics Recommendation and Tracking-iCART

Abstract

It is very common to use trolleys in supermarkets, they are machines which help us in easily carrying around a lot of items in the supermarket. iCart aims to extend the services offered by these trolleys by augmenting features such as indoor navigation, product recommendation and instantaneous reply to customer queries. For indoor navigation the RSSI values of the bluetooth modules are used to find the customers coordinates and dijkstra's algorithm is used for finding the shortest routes, for product recommendation age, gender and month of the year are passed as input parameters to a classification model and for replying to customer queries a chatbot is implemented using RASA framework. All the features mentioned will be integrated in a single LCD screen mounted on the trolley. This system not only wanes the energy spent by the customer foraging for items, but also increases the owner's profits by providing product recommendations. This model is been implemented using IoT and Machine Learning techniques to save time of customer.