Drone Surface Inspection for Hard-to-Reach Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for inspecting large objects with hard-to-access surfaces are inefficient, requiring extensive human intervention, costly equipment, and pose safety risks, with imperfect detection and localization due to operator subjectivity and the need for specialized training, and are slow, necessitating prolonged shutdowns of inspected objects.
Innovation Solution
A system utilizing a fleet of flying robots equipped with image acquisition and processing modules that autonomously inspect surfaces, reducing human operator involvement, enabling access to difficult areas without scaffolding, and providing rapid, consistent defect detection and localization through image processing and centralized management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If human operators use specific equipment such as scaffolding to access surfaces, then visual access to surfaces is improved, but installation time and cost increase significantly
Solution Approach 1:
The patent replaces mechanical inspection systems (scaffolding, lifting platforms) with an aerial robot equipped with optical sensors and image processing capabilities. The robot flies autonomously around the object, capturing images that are processed to detect defects, thereby eliminating the need for physical access equipment while maintaining inspection effectiveness.
2Ease of operation
If human operators use binoculars or optical means to access difficult surfaces, then visual access is improved, but detection effectiveness is reduced
Solution Approach 1:
The patent replaces manual optical inspection tools (binoculars) with an automated aerial robot system equipped with high-resolution cameras and sophisticated image processing algorithms. The system captures detailed images from multiple angles and uses automated defect detection algorithms, significantly improving detection precision while maintaining access to difficult surfaces.
Solution Approach 2:
The patent introduces an intermediary system consisting of the aerial robot and image processing unit that acts as a mediator between the inspector and the difficult-to-access surfaces. This intermediary captures and processes visual information, eliminating the need for operators to physically access or manually view difficult surfaces, thereby maintaining both access and detection effectiveness.
3Productivity
If multiple human operators are deployed to improve inspection speed, then productivity increases, but safety risks and costs increase
Solution Approach 1:
The patent replaces multiple human operators with a single autonomous aerial robot system that can inspect large surfaces quickly and safely. The robot flies autonomously around the object, capturing images at high speed without exposing any personnel to safety risks associated with working at heights or near large structures.
Solution Approach 2:
The patent implements a self-service inspection system where the aerial robot autonomously navigates, captures images, processes data, and generates inspection reports without human intervention during the inspection process. This eliminates the need for multiple operators while maintaining high productivity and zero safety risks.
4Adaptability or versatility
If human operators perform manual inspection, then flexibility and adaptability are maintained, but detection precision and consistency are reduced due to subjectivity
Solution Approach 1:
The patent introduces an intermediary image processing system that acts as a consistent mediator between the captured images and defect detection. The automated algorithms process all images uniformly, eliminating human subjectivity and ensuring consistent detection criteria are applied throughout the inspection, while the system remains adaptable to different object types and inspection requirements.
Data Source
Figure 1~2
Figure 3~4
AI summary
The invention relates to a system for automatically inspecting a surface of an object such as an aircraft (54), a transporting vehicle, a building or a work of art, said surface being liable to contain a defect. The system is characterized in that it comprises a fleet comprising at least one drone (14a, 14b, 14c), each drone comprising a module for acquiring images of at least one section of the surface to be inspected, and a module for processing the acquired images, which module is suitable for delivering a piece of information representative of the state of each inspected surface portion, which piece of information is called the result of the processing.