New Approach for Detecting Smartphones Securely for Disassembly Tasks

Abstract

The constantly increasing demand for raw materials is pushing the material footprint ever higher. As a result, planetary boundaries are being reached earlier and earlier each year. To counteract this trend, it is essential to reuse material and think in terms of cycles. This includes recycling activities, especially in high-wage countries. To make this economically viable, these processes need to be automated and secure. Many mobile devices contain lithium batteries, which are dangerous to humans and the environment if damaged. To recover the raw materials cost-effectively, a system is needed to remove these batteries securely. This paper presents a safe strategy to reliably identify old smartphones to provide appropriate disassembly strategies for a collection site. This enables the collection sites, a fully autonomous and, above all, safe extraction of the batteries from smartphones. This is done by using convolutional neural networks in combination with ontologies.