[5G Future] 5G promises to make smart factories even smarter
A self-driving robot automatically pulls out different materials and delivers them to an assembly station. Staff then take over, using augmented-reality (AR) glasses that show them where components need to go and link them to managers in a control center that can monitor the process and tell them what to do if things go wrong.
On the ceiling, a 360-degree camera is always watching, working in collaboration with sensors to automatically record anything that goes wrong or if a human employee goes somewhere they’re not supposed to be.
Opened in March this year, the 5G Smart Learning Factory covers 990 square meters (10,656 square feet) and is the largest training center for smart factory technology in Korea. Ranging from assembly and quality tests to packing, each of its 20 production stations has its own “smart” edge that makes use of technologies like artificial intelligence (AI), cloud computing, robotics and AR.
Since last year, Korea’s three mobile carriers have showcased 5G smart factory products and services, saying they will revolutionize Korea’s traditional manufacturing industries. But looking back, the concept of smart factories is not new. Robot arms, 3-D printing and big data analysis have already been deployed in numerous factories around the world.
Then how exactly can 5G make smart factories even smarter?
From left, KT showcases a collaborative robot arm and machine vision sensors at a press event in central Seoul. An exhibit shows how SK Telecom’s Smart Base Blocks can be used. A student at the Korea University of Technology and Education in Cheonan, South Chungcheong, learns how to use factory equipment using augmented reality glasses. [KT, SK TELECOM, KOREA UNIVERSITY OF TECHNOLOGY AND EDUCATION]
The idea of smart factories goes beyond automated equipment and robots. The “smartness” comes from the ability to solve problems in the production field using data collected from equipment.
In appearance, the difference between the smart factories that came before and after 5G may not be stark. Robots, big data and cameras all existed long before 5G shot onto the scene.
But this doesn’t mean 5G can’t bring new changes to the manufacturing field.
Professor Kim Seong-lyun of Yonsei University’s school of electrical & electronic engineering pointed out that 5G will benefit smart factories in two ways: low latency and massive connectivity.
“5G will enable low latency connectivity between different equipment and also increase how much equipment can be connected. All of the data is sent to a cloud and the scale of this will be on a level unmatched by other wireless networks,” Kim explained.
“These two factors combined will realize functions that were not possible before 5G, like immediately controlling the factory at a distance in real-time or using deep learning technology to enhance equipment efficiency.”
Using the 360-degree camera at Korea Tech as an example, 5G is what enables the camera to immediately turn toward a corner of the factory when sensors spot a problem. If the AR glasses can’t send footage in real-time with zero latency, managers won’t be able to give immediate instructions to workers on site in an emergency.
In terms of effectiveness, a more timely monitoring system can prevent breakdowns and reduce the number of faulty products. According to government survey results published in May, 30 percent of 5,003 state-approved smart factories in Korea experienced higher profitability. Forty-three percent said the rate of faulty products declined while 15.9 percent said production costs were reduced.
A report from KT Economic and Management Research Institute last year predicted that 5G can offer economic benefits to manufacturers worth 2.8 trillion won ($2.4 billion) in 2025 and 5.2 trillion won in 2030, by cutting costs.
Another effect 5G smart factories are hoped to bring is enhancing flexibility in the manufacturing process. If the machines learn to optimize production methods according to data, they will one day be able to adjust themselves according to personalized orders. A 5G report from KT Economic and Business Research Institute said that this goes in line with a paradigm shift where consumer demand is getting more diverse and orders are becoming more specific according to personal taste.
“Smart factories that cater to different consumer needs will have to automatically make adjustments to their production lines and monitoring systems for various products,” said Yoo Seung-ryeol, a professor of engineering at said Korea Tech and the head of the smart learning factory. ”This calls for more detailed technology starting from collecting consumer data and immediately designing a mold that fits their preference.”
Big business for carriers
Korea’s 5G connectivity is currently mostly limited to the center of major cities where practically no manufacturing is actually going on. This means it may be years before the networks cover the more rural locations where a lot of manufacturing actually takes place.
Nonetheless Korea’s three mobile carriers - SK Telecom, KT and LG U+ - have been showcasing 5G manufacturing technology since late last year with the hope that it would become a lucrative business-to-business model in the future. The KT report predicted that the number of smart factories in Korea would reach 20,000 in 2025 and 29,375 in 2030.
The technologies local carriers have showcased so far can be largely divided into several categories.
First is robotics. Self-driving robots that can move components around and deliver them to assembly lines are offered in some form by all three carriers. Instead of repeatedly moving the same parts on the same route, 5G-enabled robots will be able to move with more flexibility like delivering different parts as soon as purchase orders are sent in.
Collaborative robots, also known as cobots, are the big game changer that 5G brings to the world of smart factory robotics.
Traditional industrial robots repeatedly do the same job and are heavy and massive, making them dangerous to keep near people.
Collaborative robots come in the form of human arms and are able to move in a more sophisticated and flexible way. Their role is expected to increase once 5G enhances flexibility in smart factories and accordingly robots will have to respond to what is happening rather than following a series of commands.
As in cars and games, vision-based technology is another field expected to play an important role in 5G smart factories.
Faster networks will enable the real-time transmission of higher resolution videos. This will allow for special cameras like the one installed in Korea Tech and machine vision equipment on conveyor belts that detects faults.
Cameras can already spot defective products, but doing so at a higher resolution will eliminate the need for second check-ups with the human eye and enable details as small as measuring how deep a volt is screwed into a product. 5G will also connect this vision technology with the cloud to constantly exchange and update data for monitoring systems.
AR can also be used for maintenance and training. AR headsets can help train inexperienced workers on the job. Glasses can also connect workers to managers in remote headquarters or equipment manufacturers if something goes wrong. This can allow for a timely response if a production line stops, compared to the past when factory workers had to report the issue and wait for it to be fixed.
These may sound like small fixes, but in a large factory, a small delay can make for considerable losses if the entire production line gets held up.
Another function in development is the “digital twin.”
This refers to a digital 3-D map of production equipment that is identical not only in its form but also its current status. A complete digital twin is considered a smart factory goal because, if possible, managers will be able to monitor defects immediately and even sort out what’s wrong without visiting the site in person.
Using a virtual production line to find the most efficient techniques is a possibility when the 3-D mapping technology is finished.
High hopes
Private companies are not the only ones that have high expectations for smarter factories.
Last Wednesday, President Moon Jae-in pledged to continue supporting the expansion of smart factories in Korea, with the goal of expanding the number to 30,000 by 2022 and adding an additional 2,000 AI-based smart factories by 2030. It was part of the 2030 vision for Korea’s “manufacturing industry renaissance,” through which he aims to make Korea the world’s fourth-largest exporter that year, up from the current sixth.
The government has set this year’s smart factory budget at 312.5 billion won, more than double last year’s 125.6 billion won.
However, some experts say that 5G smart factories still have a long way to go. 5G-network coverage is still limited and most non-conglomerate manufacturers are relatively short of cash - most 5G-connected equipment requires a massive investment.
“There needs to be investment made in the public sector like in medicine or education, for business owners to see that 5G can make a difference,” said Korea Tech Professor Yoo. “If they are sure that 5G is beneficial and that there’s a business there, they will invest even if it comes with high stakes.”
Researcher Nam Sung-ho of the Korea Institute of Industrial Technology questions whether 5G is a must in the manufacturing sector.
“Before talking about the nationwide application of 5G smart factories, we need to ask ourselves whether 5G is really necessary - sure there are production lines that 5G would help, but we don’t need it at every production facility,” he said.
Nam added that a lot of factories that use data still do so in such a basic way that a wired network is sufficient.
“A lot of the 5G technology at the moment is about transferring data in and outside the factory. But in order to realize a true smart factory, we need connectivity among the machines, but equipment manufacturers don’t care that much about 5G. We’re missing the basics here.”
Yonsei University’s Professor Kim, on the other hand, points out that the entire factory doesn’t necessarily have to be smart.
“We should change our perspective on seeing smart factories as something massive,” he said. “I don’t think it even has to be a factory - even connection between very small pieces of equipment can fall in the category of smart factories.”
BY SONG KYOUNG-SON [song.kyoungson@joongang.co.kr]
with the Korea JoongAng Daily
To write comments, please log in to one of the accounts.
Standards Board Policy (0/250자)