Press Relase / News Release Distribution Service [@Press]

Note: This page is a machine translation of the Japanese original and is provided for reference only.
In the event of any discrepancy between this page and the original, the original shall prevail. Click here for the original text.

Born of an Industry-Academia Collaboration between Komatsu Ltd. and Doshisha University Wallpaper AI Identification Application "KABEPITA KANSAI DX AWARD 2024" Grand Prix

コマツ株式会社

Komatsu Ltd. (Head Office: Higashiosaka City, Osaka; President: Satoshi Komatsu) and Intelligent Mechanisms Laboratory, Department of Intelligent Information Engineering, Faculty of Science and Engineering, Doshisha University (Professor: Masahiro Okuda) have jointly developed the wallpaper AI identification application "KABEPITA" equipped with an "automatic texture identification program". The award ceremony was held at the Kansai Digital Month Closing Event on October 31, 2024.


Komatsu Ltd. x Doshisha University Grand Prix Award



Initiatives that received special recognition in winning the Grand Prix

(1) Komatsu Ltd. and Doshisha University collaborated with industry-academia to solve problems using image recognition AI.

(2) With the understanding and cooperation of six major material manufacturers and competing companies, Komatsu developed an "automatic texture identification program" that identifies minute differences in the surface of materials, successfully identified popular wallpaper product numbers, and made it into an application.

(3) In order to solve the identification problem and develop the application, Komatsu established the DX Solution Division within Komatsu Ltd. Ltd. to develop the application, and established a system for seamless collaboration with Doshisha University and partner companies.


Wallpaper AI Identification Application "KABEPITA



■Outline of "KABEPITA

Komatsu Ltd. is a Higashi-Osaka-based company whose main business is interior design with over 45 years of business experience.

There are more than 190,000 interior construction companies with capital of over 1 million yen throughout Japan. While the industry is facing a shortage of human resources and long working hours, there is a mountain of inefficient work that has remained unchanged for decades. One such task is the identification of materials with unknown part numbers, which is performed manually several times a month at the request of customers.

In order to solve this inefficient human work and identification problem, the Intelligent Mechanism Laboratory, Department of Intelligent Information Technology, Faculty of Science and Engineering, Doshisha University, and Komatsu Ltd. took on the challenge of solving the problem by utilizing image recognition AI through an industry-academia collaboration.


Introduction of the "KABEPITA" project at the award ceremony


The conventional identification program could not identify minute differences in material texture, and a new program had to be built. 2 years were spent developing an "automatic texture identification program," which made it possible to identify product numbers even for materials with only minute surface differences in identification factors.

Even if an efficient tool is developed based on the identification system, it is not widely used because the construction industry is dominated by elderly people who do not want to learn complicated operations or acquire specialized knowledge. In developing the application, we developed "Kabe Pita," an application that allows anyone to easily identify wallpaper part numbers with easy-to-understand buttons and intuitive operations so that the work can be completed.


Komatsu Ltd. x Doshisha University Grand Prix Winners



■Kansai Digital Month

The month of October each year is designated as a month for strengthening the promotion of DX (Digital Transformation) in the Kansai region, and a variety of initiatives are implemented through public-private partnerships.


[Kansai Digital Month Executive Committee]

Kansai Economic Federation, Kansai Economic Federation, Kinki Bureau of Telecommunications, Ministry of Internal Affairs and Communications, Kinki Bureau of Economy, Trade and Industry, Ministry of Economy, Trade and Industry, Kansai Information Center, Information-technology Promotion Agency, Organization for Small & Medium Enterprises and Regional Innovation, JAPAN

https://www.kankeiren.or.jp/kansai-dx/digital-month/



KANSAI DX AWARD 2024

The Kansai DX Award 2024 is a system to honor advanced companies in the Kansai region that are engaged in DX activities.

By widely disseminating advanced case studies of companies, including small and medium-sized enterprises (SMEs), the awards will serve as a reference for other organizations and help raise the level of DX initiatives in the Kansai region.

https://www.kankeiren.or.jp/kansai-dx/award/



Notice of "KABEPITA" Exhibition

The "KABEPITA," which won the Grand Prix in the "KANSAI DX AWARD 2024," will be on display in Osaka and Tokyo.


Osaka [Mydome Osaka]

The 37th Higashi-Osaka Industrial Exhibition "Techno Messe Higashi-Osaka 2024

Exhibition: November 6 (Wed.) - 7 (Thu.), 2024

Details: http://www.hocci.or.jp/wp/?page_id=7927


Tokyo [404 NOT FOUND (Shibuya Sakura Stage 4F)

The 1st Tech Direction Awards Winning Exhibition

Saturday, November 16 - Sunday, November 17, 2024

Details: https://award.tech-director.org/event01

We will participate in a talk event.

Date: Saturday, November 16, 14:00-15:00

    Session #2 "Co-Creation with Other Companies



■■KABEPITA Official Website

https://textorage.com/kabe-pita/



■■Company Profile

Company name: Komatsu Ltd.

Head Office: Komatsu Building, 1-15-14 Kawamata, Higashi-Osaka, Osaka, Japan

Representative Director : Satoshi Komatsu

Business : Interior design, sports, specialized construction, wholesale

Capital : 10,000,000 yen

Registered builder: Building construction business, Osaka Governor's License (Han-4) No. 53020

       Machine Tool Dealer License No. 62222R045302 issued by Osaka Prefectural Public Safety Commission

HP : https://komatsu-coltd.co.jp/

Logo Image