UDMTEK

UDMTEK has developed machine language processing technology for the first time at home and abroad
to provide a variety of application services by interpreting the execution of unknown control logic running
inside the machine and the characteristics of data flow as an explainable AI approach.
News & Trends
UDMTEK paying attention to the digital twin
2021-03-09
2688

UDMTek paying attention to the digital twin

“AI is the most experienced engineer in the factory in the future”

 

 

 


 

 

Company A's Seoul headquarters integrated monitoring center. Virtual reality (VR)-based screens are floating on dozens of monitors. The names of places such as Frankfurt in Germany, Alabama in the United States, and Zhangjiagang in China are written on each. Floating on the screen is the VR image of each overseas factory. It is a cyber factory that reproduces the actual factory in real time.

 

At that time, a red light comes on in one of the Frankfurt factory lines. There was a problem with one machine. Staff at the monitoring center analyze the information sent by the machine to quickly find out what went wrong and fix the problem. This is thanks to the artificial intelligence (AI) installed in the center system that provided solutions to similar failures in the past. 'If it was like before, you would have had to contact the machine manufacturer and send it to Frankfurt… .’ The center staff recalls the past and leaves work on time.

“This integrated management system through cyber factories is not a distant future. We are installing systems for factories that have already automated processes. We can manage our factories in Europe and America from our headquarters in Seoul.”

 

On the 3rd, CEO Wang GI Nam, who met at UDMTEK office in Suwon-si, Gyeonggi-do, said. The hypothetical case of Company A shows the technologies that can be implemented by UDMTEK's CPS (Cyber ​​Physical System). CPS refers to the fact that the cyber factory and the physical factory work together in synchronization like twins. It means that a factory that is identical to a real factory exists in cyberspace. UDMTEK is a small and medium-sized company with the technology to create future smart factories.

Global companies such as General Electric (GE) and Siemens are competitors of UDMTEK. The'digital twin' technology is gaining attention as a future technology with the spread of smart factories. Wang, who is also a professor at Ajou University's Department of Industrial Engineering, founded the company in 2007 at Ajou University's laboratory. Since its inception, it has grown in close cooperation with Hyundai-Kia Motors' production lines. Since 2015, it has begun commercialization, and has recently been supplying digital twin technology to more than 40 overseas plants of Hyundai and Kia Motors and LG Display.

CEO Wang said, “Global competition is possible because we have advanced technology by using an actual workplace with Korea's best production system as a “laboratory”.”

 

The biggest advantage of cyber factories is the efficiency that saves time and money. In general, to build a factory, engineers design, install equipment, and test-run to catch problems to create an optimized factory. However, by using a cyber factory, optimized design is possible through virtual commissioning before even building. According to UDMTEK the production stabilization period including commissioning can be shortened by more than 40%. It is also useful for plant management and maintenance. When there is a problem on the production line, you can quickly see which machine is having a problem and what is the cause. In the past, mainly by contacting the customer service center of the company that made the device, an engineer had to go directly to the site to fix the problem. During that period, the line was forced to rest.

The reason digital twin technology is considered the future technology of smart factories is that it can ultimately maximize productivity by combining with AI. CEO Wang said, “Each of the plant managers, workers, and engineers has the knowledge that they have accumulated through “know-how”. AI will learn all the factory data and become the oldest engineer in that factory.”

 

 


2018.12.11
동아일보 김현수 기자 kimhs@donga.com