Dados do Trabalho

Título

ARTIFICIAL INTELLIGENCE (AI): AN INNOVATIVE APPROACH FOR AUDITING FIELD SAMPLING IN WASTE MANAGEMENT CONTEXT

Resumo

Audit and control models are necessary in an open, transparent and accountable government. The use of Artificial Neural Networks (ANN) in audit models to predict the physical properties of Municipal Solid Waste (MSW), such as gravimetric composition, has been discussed, especially after the world experienced the COVID-19 pandemic. Traditional on-site sampling is expensive, slow and requires specialized professionals, who are exposed to physical and chemical risks, while predictions made by ANN can be carried out with little or no waste handling, and can be done retroactively, to fill gaps in information left during the mandatory quarantine period. ANNs depend on data sets provided by third parties, which need to be chosen looking for reliability, periodicity and availability. In this work we show the possibility of adopting ANN models fed by socioeconomic datasets relating to an important capital in Brazil to make predictions of the physical properties of MSW produced in this city destined for landfills.

Palavras-chave

Resíduos Sólidos Urbanos; Redes Neurais Artificiais; composição gravimétrica; aterros sanitários; pandemia

Arquivos