Autonomous Biomedical Waste Segregation Robot

Abstract

The increasing volume of biomedical waste generated in hospitals and clinical environments presents serious challenges to environmental sustainability and occupational safety. Improper segregation and manual handling of hazardous waste, particularly infectious materials and sharp objects, significantly increase the risk of disease transmission and accidental injuries among healthcare personnel. These risks are further intensified in developing regions where biomedical waste management practices remain largely manual and insufficient. Consequently, there is a growing need for intelligent and automated waste segregation solutions that can improve safety and operational efficiency. This research presents the design and implementation of an Autonomous Biomedical Waste Segregation Robot that integrates computer vision–based classification, robotic manipulation, ultraviolet sterilization, and cloud-enabled monitoring into a unified framework. The proposed system uses an ESP32 camera module to capture real-time visual data, which is processed using a YOLOv8 deep learning model trained to classify biomedical waste into infectious, sharp, and non infectious categories. Based on the classification results, a simulated Franka Emika Panda robotic arm, implemented in the PyBullet environment, performs automated pick-and-place operations, including UV sterilization prior to disposal into category-specific bins. An Internet of Things architecture connects the detection and segregation pipeline to a Firestore database and a Streamlit-based web dashboard that provides real-time system visualization, bin fill level monitoring, historical data analysis, automated alert generation, and predictive analytics. The prediction module estimates future bin fill levels using historical waste data, enabling proactive planning of waste collection activities and improving operational decision-making. The results demonstrate the technical feasibility of integrating deep learning, robotic automation, ultraviolet sterilization, and cloud-based monitoring into a single autonomous biomedical waste management system. The proposed framework offers a scalable and cost-effective approach that can be adapted to healthcare facilities of varying sizes. By reducing manual intervention and enabling predictive monitoring, the system improves safety, supports regulatory compliance, and contributes to more intelligent and reliable biomedical waste management practices