"Alexa, como vai minha plantinha?" : Sistema de monitoramento inteligente utilizando assistente virtual para auxiliar no cultivo caseiro de hortaliças

Abstract

The practice of cultivation and gardening marks history with different purposes, including improving the physical and mental health of practitioners. Cultivation of vegetables provides a healthier and more balanced diet, as it provides foods rich in iron, vitamins, and fiber. Thus, organic farming is a reality in the lives of many families, who use improvised spaces in their homes to plant. From this, different researches and market products act in the cultivation monitoring, intending to provide better results in the harvest. The development of equipment to read environmental variables and diagnose the quality of plant cultivation, with which users generally interact through traditional interfaces, such as websites and applications. In most cases, this equipment is dedicated to large plantations. In this context, this project aimed to develop and evaluate a prototype of a monitoring system for home gardens, which users can interact with through Alexa devices. From this, they can inform themselves about the care that their cultivation needs at the moment, as well as some general tips, through a request-response during the interaction with the virtual assistant. Based on the study of the main characteristics of the plants’ life cycle, as well as monitoring parameters, the work began with a survey of the requirements. From that, the system was modeled and developed. Presenting a concern with users, the work also reports an evaluation of the system that sought to understand some issues related to usability. However, it was possible to point out that the language contained in the interaction of users with the system is simple and easy to understand. In addition, it was seen that the information is extremely useful in your daily life, as it can provide practicality and greater knowledge about how to take care of your vegetables. Despite this, the need for a re-design of the interaction modeling based on a specific methodology for voice interaction was noticeable.