Related Experiment Video
Updated: Oct 7, 2025
11:15
Automated Visual Cognitive Tasks for Recording Neural Activity Using a Floor Projection Maze
Published on: February 20, 2014
13.2K
False Ceiling Deterioration Detection and Mapping Using a Deep Learning Framework and the Teleoperated Reconfigurable
Archana Semwal1, Rajesh Elara Mohan1, Lee Ming Jun Melvin1
1Engineering Product Development Pillar, Singapore University of Technology and Design (SUTD), Singapore 487372, Singapore.
Sensors (Basel, Switzerland)
|January 11, 2022
Summary
This study introduces an automated system for detecting and mapping false ceiling deterioration using a deep learning model and a robot. The framework achieved 89.53% accuracy in identifying structural defects, HVAC issues, electrical damage, and pest infestations.
Area of Science:
- Robotics and Automation
- Artificial Intelligence
- Structural Health Monitoring
Background:
- Periodic inspection of false ceilings is crucial for building and human safety.
- Current inspection methods are labor-intensive and pose risks to human inspectors.
- Common issues include structural defects, HVAC degradation, electrical damage, and pest infestations.
Purpose of the Study:
- To develop and evaluate a framework for automated false ceiling deterioration detection and mapping.
- To leverage deep learning and robotic systems for safer and more efficient inspections.
Main Methods:
- A deep-neural-network-based object detection algorithm was developed.
- A custom dataset of false ceiling deterioration images was created, covering four classes: structural defects, HVAC degradation, electrical damage, and infestation.
- The 'Falcon' teleoperated robot was utilized for data acquisition and real-time field trials.
Main Results:
- The object detection algorithm achieved an 89.53% detection accuracy in a real false-ceiling environment.
- The framework demonstrated accurate deterioration detection and mapping capabilities.
- Field trials validated the system's performance in practical scenarios.
Conclusions:
- The proposed framework offers an accurate and efficient solution for automated false ceiling inspection.
- The integration of deep learning and robotics enhances safety and reduces the labor involved in inspections.
- This technology has the potential to significantly improve building maintenance and safety protocols.
