Ubiquitous fridge with natural language interaction

Renato Ferrero, Mohammad Ghazi Vakili, Edoardo Giusto, Mauro Guerrera, Vincenzo Randazzo

Abstract: Past years have seen a dramatic increase in the development and use of connected devices to improve life quality. The introduction of smart appliances aims at addressing everyday issues in a home environment, such as reduce power consumption and waste, and automate some recurring processes. One of the main places where this ubiquitous computing revolution may take place is the home kitchen, as it requires constant attention to goods and products which may expire or run out soon. This paper analyzes the use of RFID technology as a way to enable food inventory in a smart fridge device, primarily aiming at reducing waste. By combining the use of Natural Language Interaction and product identification by means of RFID tags, the smart fridge application allows to get an instant overview of the fridge content, as well as an alert triggered by food expiring soon. Furthermore, the use of a real time cloud database integrated with the natural language agent allows to automatically check the availability of a certain product at home and determine the feasibility of a set or recipes. The application is integrated with the Google® Assistant development framework and it is released as a smart application add-on, allowing anyone who owns a smartphone to interact with the smart fridge device. Finally, the RFID reader and the microcontroller used for this application do not require a high price overhead and may be easily embedded within a newer smart fridge or integrated in a standard fridge.