Curupira Project: A Platform for Intelligent Monitoring of Waste in Amazon Rivers

Abstract

Context: River pollution is a critical socio-environmental problem that has shown exponential growth over the last few years, causing numerous global problems.Problem: The inadequate disposal of garbage in the rivers located in the Amazon region has caused the worsening of the degradation of the environment, affecting from the urban population to the riverside.Solution: A solution based on Computer Vision techniques is proposed for intelligent monitoring of the degradation of tributaries in the Amazon, where methods for detecting and quantifying the incidence of surface garbage are contemplated.IS Theory: This work was conceived under the aegis of the General Theory of Systems, in particular with regard to the interactions between the parts of a system. In this case, the parts are system-environment, input, output, process, state, hierarchy, goal-direction and information.Method: Aerial image data is captured by a drone camera and the image classification is done through Digital Processing Images and CNN algorithms. Then, degradation data is displayed on a web plataform, with analytics tools such as dashboards and heatmaps.Summary of Results: From the results, it is possible to highlight the Curupira platform, which has a geographic and temporal mapping system for the location of garbage in streams, based on a CNN network with 97% accuracy in detecting garbage in aerial images.Contributions and Impact in the IS area: The use of emerging technologies in IS combats the inappropriate disposal of waste in rivers, also helping in decision-making by stakeholders in the problem. Methods are established to deal with IS challenges from the perspective of sustainability, technologically impacting the Sustainable Development Goals in Brazil, as well as promoting IS for a more Humane World.