https://doi.org/10.1051/epjconf/202636702007
GreenSort: A Vision-Centric Robotic System for Smart Waste Collection and Environmental Protection using SSD MobileNet V2 and the TACO Dataset
1 School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, India
2 Centre for Neuro Informatics, School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, India.
* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.
Published online: 29 April 2026
Abstract
As part of sustainability, care for the environment is necessary to ensure health of everyone. Sadly, pollution and particularly stray waste have turned out to be a major problem for humans and wildlife. All the litter, which is not properly disposed, adds to the destruction of the environment. This research discusses a new solution to this problem: a trash picking robot. The proposed robot uses a visual collection system to detect and collect nearby trash, making it a great tool for multi-area cleaning. Robotics has changed our lives in many ways, saving time, money, and helping in many areas. This robot is designed to analyse the size, shape, and position of items to accurately, pick them up and transfer them to different types of waste classification. The implementation begins by identifying pollution problems and exploring existing solutions. Multiclass segregation of different types of wastes using SSD MobileNet V2 on TACO Dataset is performed. The TensorFlow model provides an accuracy of 96%. Ultimately, the goal is to show that technology plays an important role in keeping our world clean and green.
© The Authors, published by EDP Sciences, 2026
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
