Safety Protocol for Collaborative Human-Robot Recycling Tasks

Abstract

The recycling industry in Colombia needs improvements to keep up with market evolution. Currently, low-income families are the ones in charge of collecting recyclable materials directly from mixed-waste bins and taking them to public collection centers, where they are manually classified. Collection centers become an interesting scenario to introduce collaborative robots in the monotonous and unhealthy task of classifying plastic bottles. However, safety and reliance are critical points that need to be ensured. This paper presents the design of a safety protocol for the waste separation task. A system that integrates computer vision with an industrial robot is proposed to secure the human-robot interaction, based on ISO 10218-2 and ISO/TS15066 norms. The protocol assessment is conducted on a testbed created to emulate a pre-classification workstation of plastic bottles.