Visual inspection solutions
It is typical to carry out visual inspection using human vision, but there is variation in worker judgment, and workers ofter miss anomalies.
This also places a heavy burden on evaluators, and inspection speed cannot be increased.
Previous rule-based image inspection is capable of high-speed high-volume processing. However, it is extremely difficult to set parameters with an understanding of rules for all anomalous conditions, and cases of missing anomalies tend to occur easily.
Unlike previous rule-based image processing, our ADFI uses a web application to provide the latest AI technology enabling high-precision anomaly detection with easy parameter setting.
ADFI is anomaly detection and visual inspection system which operates on the cloud, and is equipped with state-of-the-art AI algorithms.
It allows you to adopt an AI anomaly detection and visual inspection system in a short time, simply by providing a few dozen images.
You can also check detection precision beforehand through a free trial.
You can select either of two algorithms (Deep Metric Learning or PCA-MSPC) as the AI to be used, whichever best suits the characteristics of the items to be inspected and detected.
DML (Deep Metric Learning) is a new deep learning technique suited to anomaly detection.
It enables anomaly detection even for problems that are difficult to handle with conventional anomaly detection technology (e.g., cases where the position of the inspected object is not fixed).
ADFI uses a unique algorithm which improves conventional DML, and no specialized knowledge is needed because optimization can be performed through automatic parameter adjustment.
PCA-MSPC (Multivariate Statistical Process Control based on Principal Component Analysis) is an anomaly detection technique currently in wide use in fields such as manufacturing industries.
This AI algorithm is suitable for situations where the position of the inspected object is fixed inside the image.
ADFI uses a unique algorithm which improves conventional PCA-MSPC, and no specialized knowledge is needed because optimization can be achieved through automatic parameter adjustment.
Service / Specifications
|Minimum number of images||Normal image||20 images||10 images|
|Anomalous image||5 images||-|
|Recommended number of images||Normal image||50 - 200 images||30 - 60 images|
|Anomalous image||10 - 50 images||Unnecessary|
|Maximum image size||No limitations (Within ADFI, images are automatically resized to 800x800 pixels or less)|
|Conditions of inspected object||No conditions||Fixed position and shape|
|Plan||Free Plan||Online Plan||Local Plan|
|AI model creation and its performance verification||✅||✅||✅|
|Run AI model in your local environment||✅(Up to 100 days)||-||✅|
|Run AI model in online (ADFI cloud) environment||✅(Up to 300 images)||✅||-|
|Time required to obtain AI model result||-||2-3 seconds / image||0.2 seconds / image|
|Usage fee||Free||0.02 US$ / image|
* Minimum fee is $6 / month
|600 US$ / month|
|Payment method||-||Credit card||Credit card|
|Available period||1 year from last login||Unlimited (automatically updated every month)||Unlimited (automatically updated every month)|
Cutting-edge AI technology
ADFI enables development of high-performance visual inspection systems with a small amount of learning data.
Furthermore, anomalies can be detected with higher inspection precision than previous visual inspection algorithms, and this reduces release of defective items and time spent on reinspection.
There is no need to procure servers or edge computers, or purchase expensive AI visual inspection software. You can quickly adopt AI visual inspection by using the ADFI web application on the cloud. Initial cost is low, and fees are set according to usage, so a system can be operated at low cost.
No need for servers
High detection rate
ADFI is provided on the AWS cloud, and thus can be used by connecting to the Internet.
Communication is done via HTTPS, so the ADFI API can be used safely.
When an image is sent to the ADFI API from a control PC or PLC in a production process, the inspection result is returned in real-time.
You can do a trial on the same day. The process up to operation can be completed in as little as 1 day.
Inspection for cloth fraying
Harness wiring color inspection
Food product breakage inspection
Inspection for presence of food products/ingredients
Inspection of component color codes
Inspection for bent or missing pins of electronic components
Inspection for label presence