Our mission is to develop an intelligent robotic system that can identify, grasp, and sort waste into recyclable and non-recyclable categories, enhancing recycling efficiency and promoting sustainability.
Effective waste management is crucial for environmental conservation. Automating the sorting process can reduce human effort, minimize errors, and increase recycling rates. This project explores the integration of machine learning for object recognition with robotic manipulation to create a practical solution for recycling challenges.
Our end goal is to design and implement a robotic system using the Sawyer robot arm that can autonomously identify, grasp, and sort various waste items into designated recyclable and non-recyclable bins.
Why This is Interesting:
Developing a robotic sorting system involves multiple interdisciplinary challenges:
Object Recognition: Leveraging advanced vision-language models to classify diverse items (e.g., tissue, paper, cans, bottles, cloth).
Robotic Manipulation: Accurately localizing and grasping objects with different shapes and textures.
Coordinate Transformation: Mapping 2D camera detections into the robot’s 3D workspace using AR tags and calibration techniques.
Smart Recycling Stations: Automated sorting in public recycling bins.
Industrial Waste Management: Streamlining waste processing in manufacturing facilities.
Residential Recycling Systems: Assisting households in sorting recyclables effectively.