Robot for automatic waste sorting on construction sites
As a large amount of waste generated from global construction activities, robots that can automatically recycle construction and demolition (C&D) waste have become efficient tools for conserving natural resources, but the complex environment and high diversity of waste on the construction site r...
Gespeichert in:
Veröffentlicht in: | Automation in construction 2022-09, Vol.141, p.104387, Article 104387 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | As a large amount of waste generated from global construction activities, robots that can automatically recycle construction and demolition (C&D) waste have become efficient tools for conserving natural resources, but the complex environment and high diversity of waste on the construction site raise challenges for robot patrolling, object recognition, and grasping. This paper describes a robot for C&D waste recycling, achieving real-time navigation through Simultaneous Localization and Mapping (SLAM). Additionally, a deep learning method and a high-precision 3D object pickup strategy were adopted for the accurate identification and stable grasping of waste items. The recognition accuracy of various kinds of C&D waste was analyzed under different illumination and spatial density conditions. Based on this research, the automation level and the application scenario of the robot prototype would be further improved and broadened.
•Built a dataset of waste objects on the construction site•Designed and built a robot prototype for automatic waste sorting on construction sites•Developed a novel simultaneous localization and mapping (SLAM) method for high efficiency and high accuracy robot localization•Proposed a 3D object grasping strategy to make the robot stably grasp the waste objects•Studied the influence of light conditions and spatial density on the recognition accuracy of waste objects |
---|---|
ISSN: | 0926-5805 1872-7891 |
DOI: | 10.1016/j.autcon.2022.104387 |