AI Robotics LTD. Begins Offering ?-version of ADFI, the World’s First Image Anomaly Detection SaaS With Automatic Parameter Optimization

Launch of AI cloud service for visual inspection, equipped with world’s first automatic parameter optimization function

In just a few hours(*), users can develop anomaly detection AI for visual inspection, and integrate it into a real-world environment

KANAGAWA, Japan–(BUSINESS WIRE)–On February 3, AI Robotics LTD., a firm providing AI cloud services, began offering a North American version of the ? version of ADFI, an image anomaly detection SaaS for visual inspection in factories.

( https://adfi.jp )

ADFI is an AI cloud service enabling automated development and implementation of image anomaly detection AI for visual inspection in factories.

Previously, development of anomaly detection AI required parameter adjustment and verification by experts to find the proper parameters of the AI model, and this typically required a development period of a few weeks.

ADFI is equipped with the world’s first function for automatically optimizing the parameters of an anomaly detection AI, and this enables development of anomaly detection AI and adoption in a real-world environment in just a few hours(*) without specialized AI technology.

Also, ADFI uses a pay-per-use system, and an AI model can be developed without initial expenditures. Therefore, adoption of AI can be accelerated in visual inspection processes in all sorts of factories.

*Time required for AI model development, implementation, and verification only

About ADFI

ADFI is an image anomaly detection SaaS for visual inspection in factories.

Anomaly detection AI for visual inspection can be easily developed in a short time without specialized AI technology.

ADFI can be tried out for two weeks at no charge from the following URL.

https://adfi.jp

Contacts

AI Robotics LTD.

Takuya Harada

TEL: 81-50-5211-8474 (Japanese correspondence)

Mail: biz-team@airobotics.jp (English and Japanese correspondence)

Please contact us by e-mail.