[Korea and the fourth industrial revolution <16-1 Smart Factories>] The future factory (some assembly required)

Aug 07,2017
The word “smart factory” might conjure the image of robots busily moving down an assembly line, sifting through parts, tightening screws and stuffing products into boxes.

But in actuality, most factories have long implemented such automated processes. What makes a smart factory truly “smart” is having machines with problem-solving capabilities that can cut defect rates, prevent accidents and even produce bespoke products.

One example of a factory that has neared full automation is Siemens’ Amberg facility in Germany. There are over a thousand sensors installed in machines across the plant that collect information on everything from product quality to machine condition.

The results are sent to a computer mainframe that can analyze the data and adjust machines’ performance accordingly. Siemens boasts that the system has helped reduce its product defect rate from 500 per million 20 years ago to just 10. Energy consumption has also declined by 30 percent, the company said.

The research firm Markets and Markets projects that the smart factory market will grow an average 9.3 percent annually from this year until 2022 to reach $205.42 billion. With smart factories proving themselves capable of dramatically increasing productivity, manufacturers of all sizes, from big conglomerates to small suppliers, are scrambling to adopt the technology in their plants.

A sensible investment

At Posco’s steel plate factory in Gwangyang, South Jeolla, as machines press large chunks of steel into sheets, sensors feed information about the production process to a piece of software that the company has developed called PosFrame.

The country’s largest steelmaker has been using PosFrame since 2015 to improve their production of metals. Every day, the sensors collect terabytes of data, and the software, equipped with analytical ability, helps managers locate defects on the line and come up with solutions.

“The use of PosFrame has cut the work of the company’s software engineers by half,” a Posco spokesperson said.

Similarly, Hanwha Techwin, a maker of aircraft engines, has attached Bluetooth sensors to each engine it creates at its plant in Changwon, South Gyeongsang. The sensors help the company track each engine through its life in the factory. Previously, barcodes have served this purpose, but they have proven cumbersome since they require constant scanning by line managers.

“The new sensors connected via Internet of Things [IoT] technology will make the final product delivery and management process more efficient and precise,” said Leem Jae-young, vice president and head of the smart factory team at Hanwha Techwin.

The company aims to have 2,000 sensors at its factory by the end of this year and 20,000 by next year, Leem said.

Hanwha Techwin has invested roughly 100 billion won ($89 million) in the factory, which began operations last November. It produces parts for global giants like General Electric and Pratt & Whitney.

UlalaLAB’s Wicon device monitors the temperature, humidity and vibration of core machines at J&H Tech’s factory in Yongin, Gyeonggi. An analysis of the data can be viewed on smartphones and tablets. [PARK SANG-MOON]
Small companies, big data

While large conglomerates are pumping huge amounts of investment into state-of-the-art smart factories, one local start-up, ulalaLAB, has focused on the smaller fry. It has been providing IoT sensors and cloud-based smart factory software to small and medium-sized manufacturers.

The idea is that smaller businesses can implement the system on top of their existing infrastructure. It may not be as fancy as PosFrame or the expansive network of sensors at Hanwha Techwin, but it does the job. The company’s system, called Wimfactory, lets small companies digitize their factories on a low budget.

J&H Tech, a manufacturer of set-top boxes and car audio systems based in Yongin, Gyeonggi, has adopted ulalaLAB’s platform. Although its plant might not give the air of a sophisticated factory - no robotic arms or cranes here - there are sensors that check the temperature, humidity and vibration levels of core production machines.

These sensors send information to a router-like device the company has created called Wicon. Wicon can take information from up to four sensors at a time and deliver the data over the internet to Wim-X, ulala’s cloud-based factory management software. The software compiles the information into a dashboard that employees can view on their smartphones and tablets to track machine conditions in real time.

“Our device makes the transition to smart factories much easier and cheaper, at about a tenth of the prices of other smart factory platforms, because the factories don’t need to change their existing settings,” said Severus Kim, chief marketing officer at ulalaLAB.

J&H Tech has been using ulalaLAB’s system for about three months, and workers say they sense the time-saving it promises.

“It has become much more convenient for us to check whether our machines are functioning well,” said Park Lee-kweon, J&H Tech’s head of manufacturing, “and since it automatically translates data into meaningful dashboards, we can easily make reports to show our customers we are performing well.”

Before J&H Tech, ulalaLAB tested its smart factory system at a parts supplier for Hyundai Motor in November 2015 under a government-sponsored project. After three months of using ulalaLAB’s system, the supplier said its defect rate went from 18 percent to below 10 percent. According to ulalaLAB’s data, the reduction has saved the factory about 20 million won per month.

Currently, the start-up’s system is being used by Nike at its shoe manufacturing facility in Indonesia and at the factories of industrial tool maker Atlas Copco Korea.

Managing machines

It would be amiss to talk about smart factories without mentioning robots. Industrial robots have been used in factories since 1959, when General Motors first employed a robot at its assembly line in New Jersey.

In the following decades, robots have been one of the most consequential pieces of technology in the demise of blue-collar manufacturing jobs. Their ability to perform repetitive labor has largely stripped the need for human workers.

Still, the factory machines of today still require some human management because they are only programmed to perform specific tasks. That may change with the advent of artificial intelligence and high-speed wireless networks that allow robots to communicate with each other and more easily adapt to different tasks.

For instance, a two-handed robotic arm developed by Swiss industrial robot maker ABB can directly learn actions from engineers by mimicking their gestures. Robots like this one that perform human-like motions offer great potential for factories because they can be installed without companies having to redesign their production lines. They can work hand-in-hand with human workers, which is why they have been called collaborative robots, or “cobots” for short. Universal Robots, Rethink Robotics, Kuka Robotics and ABB are major producers of cobots.

Although Korea has yet to produce a major manufacturer of cobots, Hanwha Techwin recently jumped into the business and released its first collaborative robot, called HCR-5, in March.

Although Hanwha is still at a very early stage in the business, the company believes robotics will be a great revenue driver in the future, especially with the growing need for robots in smart factories. According to Markets and Markets, the global cobot market is expected to expand an average 60 percent annually to reach $3.3 billion by 2022.

Bespoke products

At the end of 2015, sportswear giant Adidas opened a high-tech shoe making facility in Ansbach, near its headquarters in Germany. The factory’s big selling point: its 3-D printers that can make custom shoes.

“With Speedfactory, we are challenging conventions and disrupting the status quo of our industry,” Gerd Manz, vice president of technology innovation at Adidas, said last year.

The factory plans to produce 500,000 pairs of shoes within this year, a small number compared to the more than 300 million pairs the company produces annually. But if this factory system becomes the new normal, it could mean the death of mass production and higher demand for bespoke products.

The technology for 3-D printing still has limits, notably in speed and materials available for printing, but it could be a game changer if machines can make products faster and with more variety.

“It’s a matter of time,” said Park Han-ku, head of a local smart factory consulting firm Smart Machine & Factory and a director in the Korea Industry 4.0 Association. “Already, 3-D printers are starting to be commercialized for use in industries like aviation, automobiles and medical devices.”

General Electric is already producing fuel nozzles using 3-D printers. Another U.S.-based company, Local Motors, can print out a car body and chassis for electric vehicles in less than 44 hours.

In Korea, 3-D printer adoption is relatively slow. According to data from the Ministry of Trade, Industry and Energy released in March, Korea’s 3-D printing market has grown from 181.5 billion won in 2014 to 223 billion won in 2015. It estimates average annual growth to be an average 22.9 percent to reach 508.2 billion won by 2019.

“Market growth in Korea will lag behind global market growth for 3-D printers due to lacking demand from the nation’s manufacturing sector,” the ministry said.

Challenges ahead

According to data from the Institute of International Trade under the Korea International Trade Association, the country’s sensor technology is only at 30 percent compared to leading companies in the United States and Germany.

“Korea has relatively well-established manufacturing equipment and internet network technology, but it lags behind greatly in terms of sensor technology and radio-frequency identification,” said Lee Yu-jin, a researcher at the Institute for International Trade. “While government has been pushing to expand the number of smart factories in the country, most of the newly established smart factories remain at a very basic level.”

In 2015, the government established a public-private partnership dedicated to supporting the nation’s small and medium-sized businesses so that they can digitize their factories. According to the partnership, the Korea Smart Factory Foundation, it supported roughly 2,800 SMEs to partly digitize their factories as of last year. The foundation also laid out goals to make the accumulated figure 5,000 by the end of this year and 10,000 by the end of 2020.

However, 79.1 percent of the supported factories turned out to be at the most basic stage of smart factories, or level 1 among the four levels created by the foundation. At level 1, computerized systems can only track and document the transformation of raw materials to finished goods.

“You could call it a smart factory, but computerizing the manufacturing process and merely automating processes at factories are agendas of the third industrial revolution rather than the fourth,” said Park Han-ku of the Korea Industry 4.0 Foundation. “Rather than showing how many factories have turned into smart factories by applying the simplest systems to many factories, government should make a focus group that can be good models for other companies like Adidas’ Speedfactory.”

Lee of the Institute for International Trade said the approach should be more about quality than quantity. “The Korean government needs to focus on making manufacturing smarter rather than increasing the number of smart factories.”

BY KIM JEE-HEE [kim.jeehee@joongang.co.kr]

dictionary dictionary | 프린트 메일로보내기 내블로그에 저장