[‘Engineering Men’ Startup 4 People’s ‘Direct Speech’] “Manufacturing Competitiveness is Being Captured by China, Smart Factory is Not an Option but a Must”
"Samsung and SK have been preparing for 10 years, small and medium-sized enterprises have not yet felt the effects of labor costs and efficiency”
The goal of “Made in China 2025” is to enhance manufacturing competitiveness. It aims to introduce artificial intelligence (AI)-based smart factories into the manufacturing industry to offset the ever-increasing labor costs while producing high-quality and uniform products.
Traditional manufacturing powerhouses such as Germany and Japan have also entered the competition for smart factories. This is in response to the efficiency improvements in China's manufacturing sector. Moreover, by effectively controlling labor costs through smart factories, it is possible to easily repel challenges from emerging industrial countries such as China, Vietnam, and Indonesia. This implies that the foreign direct investment (FDI), which has been concentrated in emerging countries so far, might depend on how advanced the smart factory technologies and corporate collaboration structures are in the future.
Korea also needs to respond quickly as manufacturing accounts for 44.5% (as of 2015) of its economy. Smart factories encompass a wide range of fields, including networkization of processes and production automation based on AI and 5th generation (5G) mobile communications, control, safety, and quality management. However, there is resistance to changing production methods and systems domestically due to the clear structure of primary and subcontracting, making it a challenging environment for the smart factory startup ecosystem to flourish.
In this context, Bluepoint Partners met with representatives from four companies it is nurturing, all based on smart factories and AI, to hear their hardships, technology trends, business visions, and experiences of collaboration with accelerators in the domestic industrial field. The startups present included Park Sung-jae from Excelo, which developed an intelligent refractory integrated management system; Kim Ki-duk from CN Tech, which developed an unmanned remote management solution based on collateral; Ahn Hyun-soo from Jeep Korea, which creates solutions for detecting harmful chemical leaks; and Park Oi-jin from Acryl, which develops interactive chatbots. They agreed that Korea needs an across-the-board adoption of smart factories to enhance manufacturing competitiveness. They also urged that technology companies need stable investments for at least seven years since developing technology takes a lot of time.
Kim Yu-kyung (Moderator): What was the opportunity to meet with Bluepoint, and what was the collaboration about?
Park Sung-jae: We are applying new technologies such as Internet of Things (IoT) to refractories that receive molten metal and other hot materials. I wanted to objectively validate this item, so I applied for the ‘private investment-led technology startup support (TIPS)’ program and chose Bluepoint as the accelerator because I judged that their direction aligned with ours. Startups lack experience in attracting investment or building product identities, and receiving various opinions helped reduce our risks. The government’s startup support funds were also a great help.
Kim Ki-duk: Bluepoint felt like the chairman came in and took over management. In fact, startup founders often lack knowledge about management, even if they hold the title of CEO. If they are left to manage in their own style, they will eventually stumble or walk barefoot. They help build roles and frameworks together as a CEO. They teach how to catch fish and provide support until one can walk on their own.
Ahn Hyun-soo: Manufacturing companies can be blunt. They live in a world that is the complete opposite of IT startups. I met Bluepoint for the first time after participating in the Industrial Bank contest in 2017 and got to know a new world. Until then, I solely focused on research and development (R&D) and had no idea about attracting investment, evaluating corporate value, or how to highlight the value of technology, and I learned a lot and grew from this experience.
Moderator: Isn’t it difficult to increase added value in technology startups based on B2B business?
Park Sung-jae: The refractory business operates in a limited and closed environment like steel plants, making it extremely challenging to change production processes by introducing new technologies. Even if a genius develops innovative technology, it is hard to apply. Large companies do not want to take risks, so there is even no chance to experiment with new technologies. Without a solid cooperative relationship built over time, it is hard to overcome high barriers to entry.
“Samsung Electronics S-Line Smart Factory Paragon”
Kim Ki-duk: It is hard to feel the return on investment (ROI) and productivity increases when a smart factory is implemented. The government is investing hundreds of billions, but local personnel cannot even consider introducing it for free. While there is demand for large, specialized factories, it is very limited. It took Germany a lot of cost and time to universalize smart factories. Factories that can afford to wait are those with the capacity to digitalize their operations in 5 to 10 years. The best examples of successful smart factories are semiconductor factories (fabs). Samsung's S-Line is the crystallization of smart factories. One worker is enough. Its productivity is overwhelmingly higher than the line of the Giheung campus which is over a decade old. The effects are clear, but the majority do not recognize it, making it hard to open the market.
Ahn Hyun-soo: In factory safety management, regulations have been strengthened to prevent work orders in hazardous areas, increasing the need for automation. There is growing public interest in safety, making it align relatively well compared to other fields. However, there is a tendency to concentrate on long-established companies, and startups face challenges in trying new tests and entering the market.
Moderator: How did you come up with the commercialization idea for your technology?
Park Sung-jae: Technology is content. We need to respond to various environments with flexible thinking. Refractories are just stones, but they have to withstand temperatures of 1600 degrees and 2000 tons of pressure. Therefore, nobody thought to attach sensors, and I approached this with a light heart. Everyone talks about smart factories, but each industry has its own characteristics and different technologies must be applied. The most significant content in a steel mill is molten metal. I noted that there was no device to control this molten metal.
Moderator: Is there no generalized model for smart factories?
Kim Ki-duk: There are various types of factories in agriculture, food, and more, each with different objectives such as productivity improvement and safety. Some companies target specialized industries, while others, like CN Tech, aim for general smart factories. It’s about bringing the equipment from factories with standard processes into computers. If data and operational status are to be understood, sensors are needed, but most equipment in small and medium-sized enterprises does not come equipped with them. Therefore, they are packaging and selling sensors and data acquisition separately.
“Customization Trend Over Smart Factory Platforms”
Moderator: Isn’t there potential for export to aging manufacturing countries like Japan?
Ahn Hyun-soo: We aimed at overseas markets from the beginning. We achieved first place in the incubating challenge program of Polverse, which has the most technology in the world. Although it’s not high-tech, it is simple and accurate, and we received an evaluation that once we reach the product stage, it could be utilized, leading to technology validation and exclusive supply contracts. We are looking at the markets in the following order of importance: Korea, Japan, the USA, China, Europe, and India, and targeting steel mills, glass factories, and chemical companies. Notably, we are advancing into companies in the Tianjin region of China, which have experienced factory explosions, and into companies like Intel in the US semiconductor industry.
Park Sung-jae: B2B companies do not place much weight on publicity, but Excelo has plans for global expansion, so we have created a character and a comic book-style brochure. We plan to promote the product's functions and performance through a character to overseas buyers. There are no known cases globally of extracting data from refractories. We aim to become a unique and enjoyable cultural company, breaking away from a closed culture.
Moderator: As the US and Germany are establishing smart factory platforms, won’t this lead to a phenomenon of centralization?
Kim Ki-duk: Global research institutions are consolidating smart factories and big data, which sounds good in theory but is often detached from reality. A company that makes mice and a company that makes phone cases need to develop separate smart factory solutions. Even within the same industry, each company's style produces different data, limiting possibilities for business modeling. Results from common analytical platforms used by companies and research institutions have usually not been favorable. Thus, recently about 50% use common analytical platforms, while the other half adjust according to their specific conditions.
“There is also a project to develop AI that learns artisanal recipes”
Moderator: How competitive are Korean manufacturing companies globally?
Park Sung-jae: As for POSCO, it has grown to be on par with Nippon Steel. China’s rapid usage globally has caused their technology to develop to a nearly comparable level. In contrast to Korea where most ceramic majors gravitate toward semiconductors, many in China are entering the refractory field. In the future, China's technological competitiveness will significantly develop.
Kim Ki-duk: The quality of Korean products is better than China's but not as good as that of Germany or Japan. The price competitiveness is about in the middle. However, the likelihood of falling behind China will increase over time. Ultimately, we need to move towards a structure where humans do not interfere. Machines must take over and perform quality control affordably and accurately. The government’s policy direction is not wrong. In anticipation of this, conglomerates have already been driving digitalization. Artisans are placed at every juncture of the machinery to control recipes, and since they are irreplaceable, companies are highly dependent on them. Therefore, conglomerates are attempting to platformize the artisans’ know-how through big data.
Ahn Hyun-soo: Korea is currently at about the 1.5 flank level in the global market. All key sensors are imported and must be programmed and then automated. We need tracking data, but we cannot make core sensors and have to buy from Germany, etc. After the regulations on chemical compounds were strengthened, many companies have started developing sensors, and some products have begun to be exported from Korea. Recently, while a lot has gone to China, Korea has accumulated competitiveness to enter overseas markets in areas such as material-reactive sensors.
Moderator: What are your thoughts on the government's concentration of initial startup funds?
Ahn Hyun-soo: These days, startups are becoming a resume for employment. I understand that many young entrepreneurs see startup as a way to get a job at a company they want to join. The government also needs to focus and concentrate, as it’s essential to grow companies that can succeed. The emergence of good companies will elevate the industry as a whole. If funding is easily available, founders may become complacent, at which point they cease being startups.
“Urges against indiscriminate funding support, strengthen verification and competition systems”
Kim Ki-duk: While I agree with the intention of supporting many startups to produce success stories in Korea, the screening must be strict. Funding must be prevented from being mismanaged, as it often happens where the purpose is to check business models and founders closely, similar to the practices in Silicon Valley in the US.
Park Sung-jae: By implementing a verification system and a competitive system internally, we can improve the efficiency of fund disbursement, leading to the emergence of better startups.
Moderator: How do we prevent those who are hunting for government support funds rather than focusing on entrepreneurship?
Park Oi-jin: If someone has failed in entrepreneurship but continues to receive government support, that is also a skill. That person may have pondered various ideas and written proposals and gone through verification. It can even serve as motivation for other founders. It is better to generously provide funds for startup capital. Korea is still in the phase of learning how to nurture startups effectively. There is an urgent need for faster infrastructural support, such as legal and tax consulting services, which are currently hard to access while companies like WeWork and Fast Five spend a lot on interior design. Money should be spent on commonly needed infrastructure, and there must be efforts to systematize it so that the money does not evaporate.
Moderator: What support do mid-stage startups require?
Park Oi-jin: Practical assistance, like including startup products in government procurement or prioritizing government buying, is necessary. Reducing information asymmetries like demand-supply matching is also a solution. In terms of infrastructure, it’s essential to reduce volatile aspects such as consulting histories or data. Considerations on how to secure funding later on in a "death valley" situation (when funding dries up 3 to 7 years after starting a business) are crucial. Accelerator or venture capitalists (VC) all say they encourage founders' dreams, but when it comes to investing, they prioritize sales.
“Investors press for performance every quarter; they should trust us to wait for seven years”
Moderator: Are there any expectations from government or investors as a tech startup?
Ahn Hyun-soo: Many VCs are conservative, leading to doubts on whether they can validate tech companies. Sometimes, they fall asleep during pitches by startups. There are times I have doubts about the investors’ verification capabilities. VCs and similar investors also need guidance from the government.
Park Oi-jin: VCs often find it hard to know the technologies of their portfolio companies. They rely mainly on financial perspectives, their intuition, experience, and surrounding advice for judgment. VCs demand performance data quarterly, and if performance falls, they request substantiation materials. This puts a huge burden on startups aiming for long-term visions. Funds that can create unnecessary pressure on the growth of startups should not exist in this space. Conflicts may also arise between investors and portfolio companies. There needs to be recognition that failure is acceptable while allowing enough freedom of business.
Kim Ki-duk: More support is needed for existing startups. Many startups fail to cross the death valley and go bankrupt; it’s crucial that policies revolve around easing the financial pressure on these companies. Recently, there has been an influx of capital that prefers quick returns on investment in the startup field, leading to greater pressure. It’s crucial that capital should be willing to wait for at least seven years, regardless of a startup's revenue, to allow for stress-free business operations.
Moderator: Didn’t the government state it would establish scale-up funds?
Park Oi-jin: It's too late. Even though there has been a lot of talk about it from the private sector, it has been less than two years since a sizable fund has been formed. Even then, the primary goal seems to be capital gains rather than long-term nurturing of startups.
- Kim Yu-kyung, reporter, neo3@joongang.co.kr
