Abstract
Fall accidents are the leading cause of fatalities in the construction industry, and can occur due to various environmental hazards, such as unprotected walkways, slippery surfaces, exposed edges, and so forth. To mitigate the risk of fall accidents in construction workplaces, it is crucial to identify and locate potential fall hazards. Because conventional safety monitoring methods have been inefficient, more-effective inspection methods are needed. This study presents a cost-effective multiagent robotic system that can automatically detect and localize potential fall hazards on construction jobsites. This study focused mainly on same-level fall hazards and considered all the slipping, tripping, and falling hazards in the indoor construction environment to be potential fall hazards. The proposed collaborative robots are assembled using five low-cost hardware modules and successfully can detect and localize same-level fall hazards by integrating simultaneous localization and mapping, path planning, and computer vision techniques. The proposed affordable robotic system allows for the widespread adoption of proactive fall accident prevention methods, which can contribute significantly to the safety management of construction workplaces.
Original language | English (US) |
---|---|
Article number | 04022042 |
Journal | Journal of Computing in Civil Engineering |
Volume | 37 |
Issue number | 1 |
DOIs | |
State | Published - Jan 1 2023 |
Externally published | Yes |
Keywords
- Construction robot
- Construction safety monitoring
- Deep convolutional neural network
- Fall hazard detection
- Hector simultaneous localization and mapping (SLAM)
- Multiagent collaboration
ASJC Scopus subject areas
- Civil and Structural Engineering
- Computer Science Applications