EchoSort: Intelligent Waste Management Through Acoustic Detection

Abstract

Effective recycling requires sorting at the point of disposal, yet existing systems fail to combine multi-class material identification, physical sorting actuation, and low hardware cost within a single unit while remaining sensitive to real-world acoustic noise. This paper presents EchoSort, a passive acoustic sorting system that couples a noise-isolated chamber, a classifier, and an electromechanical actuation stage to sort waste at the point of disposal. The PC-assisted prototype achieves high accuracy across multiple materials at a low hardware cost, outperforming prior systems at a fraction of their cost. Results are discussed in relation to the UN Sustainable Development Goals, and the system’s dependence on external inference is identified as the focus of future work.