Press Release

 <Computermate provided an AI solution that can prevent defects by suggesting equipment condition values ​​to the production line of Kyungchang Industry, an auto parts manufacturer. This solution enables equipment production based on AI data rather than the experience of workers, and can reduce defect rates by preventing errors caused by human work. The photo shows Kyungchang Industry's main product, a gear actuator. / Provided by Kyungchang Industry>

The 'AI Convergence Regional Specialized Industry Support Project' of the Ministry of Science and ICT and the National IT Industry Promotion Agency (NIPA) has concluded its two-year journey this year. Six regions have been selected: Daegu Transportation Equipment and Machinery Materials and Components, Chungnam Eco-friendly Mobility, Gyeongnam Automobile Parts, Daejeon Digital Material Industry, Gwangju Medical and Healthcare, and Jeju Green Energy. This is an ambitious project that aims to consolidate regional specialized industries into future industries by converging AI technology, solidifying the regional economy, which is the root of the Korean economy, and establishing a foundation for digital transformation. We will examine how regional specialized industries are changing through the convergence of AI and introduce the results over six parts. 〈Editor's Note〉

The Daegu City and Daegu Technopark Consortium secured KRW 3.73 billion in national funds (total project cost of KRW 6.75 billion) in the transportation equipment and machine material and components sector from the 'AI Convergence Regional Specialized Industry Support Project' of the Ministry of Science and ICT and NIPA, which invested a total of KRW 34.1 billion.

Local transportation equipment and machine parts companies that participated in the business as demand companies are showing anticipation, saying that this will be an opportunity to build a foundation for sustainable development based on new technologies. Supply companies that developed AI technology have developed AI services using actual industrial site data that is usually difficult to access, opening up new possibilities and also obtaining an opportunity to enter the early business that connects industry and AI. It is a win-win strategy for demand and supply companies.

Kang Dae-ik, head of the Daegu Techno Park ICT Industrial Headquarters, said, “We plan to support not only auto parts companies but also various local companies to utilize AI technology to increase the added value of their existing businesses and continue linked businesses so that productivity can be improved.”

<Daegu Transportation Equipment and Machinery Material Parts AI Convergence Support Project>

◇ Computer Mate develops AI solution to minimize defects in actuator line at chassis factory

Computer Mate (CEOs Sang-in Seo and Sung-ho Kim) developed an AI solution that analyzes the correlation between data and defect rates and uses AI to learn optimal conditions to minimize defect rates. Gyeongchang Industrial (CEO Son Il-ho) used this to successfully minimize defects occurring in the actuator line of the chassis factory.

Computer Mate collects and monitors conditional information on the operation status of actuator equipment in preparation for sudden breakdowns or defects in actuator production/manufacturing equipment and inspection equipment. This allowed preventive inspection and preemptive measures to be taken. Not only can it reduce worker dependence, but it also reduces defects that occur in the manufacturing process. Gyeongchang Industrial expects to increase productivity by 7% per year through rapid response to abnormal situations and reduction of rework.

◇ InterX, AI-based CNC machining defect and tool life prediction system

Innotech (CEO Lee Hee-bang) produces electric vehicle reducers, automatic transmission shafts, drive unit hubs, and ball lamps. The biggest problem was the loss caused by malfunction or defect of the computer numerical controller (CNC), which is the core of product production. During the production process, processing parameter values were different depending on the skill level of the CNC operator, and in this case, defective product dimensions could not be avoided.

InterX (CEO Park Jeong-yoon) stepped forward as Innotech’s ‘AI solver’. By understanding the lifespan and replacement period of all tools, including CNC, in advance, we sought ways to significantly improve product defects and productivity. Innotech, which introduced the AI-based CNC machining defect prediction and tool life prediction solution developed by Interx, predicts that the defect rate of its main product, ball lamps, will decrease from 3.2% to 2.8%, increasing production efficiency from 80% to 90%. He said it would be possible.

◇ UDM Tech, AI-based automatic control system for riveting process

PHA (CEO Seung-hyun Heo) is a company that develops and produces door latches, door modules, and door hinges. In the riveting process that uses metal rivets to connect parts, when changes in setting values ​​are required due to changes in vehicle models or process environments, these were manually entered. This often resulted in process delays or errors.

PHA joined hands with UDM Tech, which will develop AI technology for this project. The two companies collaborated on the development of AI-based quality prediction and servo axis position control systems for AI and big data
-based riveting processing processes. Applying the AI-based automatic control system for the riveting process developed by UDM Tech will allow for immediate response to subtle process changes. It can reduce delays caused by changes in setting values. It also recommends optimal setting values ​​for high quality, thereby reducing the occurrence of potential defects. It detects abnormalities in the load ratio pattern of AI-based riveting devices and calculates statistical optimal setting values, thereby reducing errors and delays. PHA announced that the production volume of actual related products per hour increased by 7% from 155 to 166, and the overall efficiency of the facility increased.

◇Seongwon Information Technology, builds a data collection and analysis dashboard

Sangsin Brake (CEO Kim Hyo-il and Park Se-jong) is a company that produces brake pads. Brake pad friction materials vary in combination and ratio depending on performance, so research and analysis takes a considerable amount of time. Brake pads require analysis of numerous performance data and linkages between items in order to design new friction material products, but it is not easy to analyze all of this data with human resources alone. Sangsin Break's long-cherished goal was to analyze data to make the process and supply chain management necessary for product design intelligent, and to make the mixing recipe intelligent during the product development process.

To this end, two companies, Sungwon Information Technology (CEO Song Seong-ho) and Big Wave AI (CEO Hee-jun Lee), collaborated to develop two AI models: 'AI-based optimal manufacturing process system' and 'AI mixing recipe intelligent system'.

Sungwon Information Technology collected and processed existing data and built a separate database (DB) for data analysis. Based on this, a performance simulator and a dashboard capable of analysis visualization were created during the compounding process. Sungwon Information Technology also developed a system design and UI/ UX linked to the existing system so that production site practitioners can conveniently utilize it.

◇BigWave AI, AI-based optimal manufacturing process construction

BigWave AI learned the DB built by Sungwon Information Technology and accelerated the development of an 'AI-based optimal manufacturing process system' and an 'AI compounding recipe intelligence system' that minimizes resources required for brake pad compounding design. In order to increase the convenience of inventory management by knowing the order lead time according to the expected production volume in the production process in advance, an order and UPH (Unit Per Hour) prediction model was developed. Based on this, Sangsin Brake built an automatic material requirements planning (MRP) system capable of more accurate predictions. In addition, we have developed an AI mixing recipe intelligence model that learns the mixing ratio of raw materials and the manufacturing process during the product development process to predict the properties of brake friction materials. We are focusing on development with the goal of increasing the mixing design hit rate and the development product process capability achievement rate.

Sangsin Brake has been able to further enhance its technological competitiveness by shortening product design time and significantly improving quality thanks to two AI solutions developed in collaboration between Sungwon Information Technology and BigWave AI.

[Source] Electronic Newspaper, Reporter Jisung Park, jisung@etnews.com