Breaking Away From Visual Inspection
What Is the AI Visual Inspection System “FALCON”?
The IoT-compatible appearance inspection system “FALCON” utilizes image processing technology to identify and detect abnormalities in the visual appearance of products, a task traditionally performed by humans. We take into account customer requirements, such as high-definition inspection, takt time support, and over-detection prevention, and offer comprehensive solutions, including the setup of the imaging environment, configuration of inspection parameters, and support for automation within the production line.
Issues and Effects in Visual Inspection
Current Issues Visually Observed
- Variations in testing accuracy due to the inspector’s skill and physical condition
- Mistakes and typos caused by handwriting of test results
AI Visual Inspection With “FALCON”
- Always make accurate judgments using set thresholds
- No input/automatic saving of judgment results
Features of the IoT Compatible AI Visual Inspection System “FALCON”
We present the steps to consider when contemplating the implementation of an appearance inspection system.
Additionally, we outline our company’s measures for addressing potential challenges at each step.
１．Providing a Visible Imaging Environment
There are so many combinations of cameras, lenses, and lighting that it’s hard to know the best configuration
By establishing a database of past inspection data, we configure an optimal imaging environment based on factors such as shape, surface condition, and details of abnormalities
２．Integration of Rule-Based and AI-Based Inspection Algorithms
Inspection using either rule-based or AI (depending on inspection manufacturer)
Optimizing inspections based on defect content through a combination of rule-based and AI approaches
３．Easy AI Inspection Setup With “FALCON”
＜Acquisition of Training Images＞
Hundreds of images required
Inspection possible with at least 10 sheets
＜Annotation (Defective Location Specification)＞
In most cases, each image takes several tens of seconds
Possible in a few seconds per sheet
＜Learning＞
Half day to 1 day
Complete in as little as 10 minutes
＜Inspection Accuracy Check＞
Confirmation is required by passing the actual inspection object through an inspection machine.
Enables bulk verification using images of the inspection target acquired in advance
４．Assistance With an All-In-One Automated System
Consultation with inspection machine manufacturer/equipment manufacturer is required.
Consistent support all in one
５．In-House Operation After Introduction
Re-adjustment of inspection conditions due to product type switching, etc. requires a request from the inspection machine manufacturer.
Easy-to-use inspection software allows you to adjust inspection conditions by following the steps. *You can also select our operational support.
IoT Compatible AI Visual Inspection “FALCON” Performance
| Take pictures of defects | Camera：Choose from area sensor and line scan camera Light：Choose from various types of lighting Imaging conditions are selected from the database |
| Learn defects | Choose from rule-based or AI AI achieves both improved accuracy and shorter learning through supervised learning |
| Inspect the product | We provide inspection equipment that works in conjunction with automatic machines designed to fit your production line. Expansion of inspection equipment according to takt time |
| Algorithm to apply | Supervised AI, Rule-based |
| Output the inspection results | Choose from JPEG, CSV, or result analysis software |
| Minimum study time | 10 minutes |
| Data storage format | JPEG、CSV |
Benefits of using IoT Compatible AI Visual Inspection System “FALCON”
Improving Inspection/Work Quality With Digitalization
■ Due to unified inspection standards by “FALCON”…
- Preventing defective products from leaking outside
- Reduction of human inspection variations
- Reduction of inspection variability within the same person
Labor Saving
■ Due to freedom from visual inspection by inspectors…
- Achieving labor savings in inspection operations
- No need to train or hire inspectors
Preventing Tampering by Acquiring Digital Data
■ Through digital data management of test results…
- Achieving traceability of inspection data
- By analyzing defective data, we can pursue the cause of defectiveness and make improvements.
Click here if you would like to learn more about the IoT compatible AI visual inspection system “FALCON”.