GET CONNECTED, a three-step process, is presented, and ecosystem structure is key.

The commonly referred to industrial IoT encompasses the concept of the Industrial Internet, which emerged as a counterpoint to the consumer Internet. The Internet of Things encompassed connections between computers and people, but this concept alone couldn't explain everything needed in industry. Ultimately, the question of how to connect machines and people led to the Industrial Internet. GE (General Electric) established the concept of the Industrial Internet early on, transforming it into a tangible solution as it evolved into a digital enterprise. Kim Sung-jin, Managing Director of GE Digital, emphasizes, "Getting started now is the first step toward the Industrial IoT." We heard from Managing Director Kim, who emphasizes that the connection must extend beyond devices to include people, to discuss GE's Industrial Internet.

Implementing industrial IoT has become an inevitable choice for manufacturing. What should be done and how?

Let me explain GE's three-step approach to industrial IoT: GET CONNECTED, GET INSIGHTS, and GET OPTIMIZED. "GET CONNECTED" involves first connecting all machines and equipment, then digitizing and retrieving data. Surprisingly, many companies haven't even achieved this yet. Next is "GET INSIGHTS," which involves analysis and prediction. Based on the collected data, they conduct enhanced analysis and forecast the future. The final step is called "GET OPTIMIZED," or optimization. This is where you gather, analyze, and predict data, and then take action to generate results that will improve your business.

-If you could give specific examples for each stage.

In reality, "GET CONNECTED" often doesn't work well. While large companies are doing well, most other companies have individual facilities, with few connecting and operating them. Furthermore, aging facilities make it difficult to extract data, hindering connectivity. "GET INSIGHTS" can be applied to chemical companies, for example. Chemical companies, where facilities are crucial, begin by properly monitoring data from those facilities. This allows them to analyze processes and predict when each facility might fail.

Building an industrial IoT system isn't something a single company can accomplish. That's why building an ecosystem is so crucial.

Yes. Especially in IoT, the ecosystem is paramount. It's not something a single company can handle. For example, in some cases, data can be collected from PLCs or controllers, while in others, sensors must be directly installed. This requires sensor companies, and SI companies to actually connect. Furthermore, for large or remote areas, a connectivity partner is crucial. GE collaborates with sensor and SI companies, has partnerships with global telecommunications companies, and has also partnered with global consulting firms. It's safe to say that GE has partnerships across all sectors.

-How did GE leverage partnerships in the cases it built?

Even when GE was working on a project last year, it wasn't done alone. There were companies that built the system and companies that supplied the sensors. This also included domestic partners. For example, some people think that GE's industrial IoT platform, Predix, is all there is to the industrial Internet, but that's not the case. For example, much needs to be done at the device level, or edge level, and manufacturing sites often require on-premise databases. Not all data needs to be uploaded to the cloud. There are partners that build industrial databases, and there are domestic companies that develop Predix apps.

"You have to start with a tangible goal
"It doesn't end with hesitation in the introduction review"

I agree that we need to build industrial IoT, but it doesn't seem to be spreading quickly in Korea. What do you think is the reason?

The most important reason is that we don't seem to understand the purpose of introducing the Industrial Internet. The Fourth Industrial Revolution is a time for innovation, and everyone believes it's an opportunity for innovation. The problem is that because they're being told to embrace it, they start exploring related technologies. But they start to question what they'll gain from it. This leads to hesitation. For example, they might say, "I need a new business model, but my current database won't allow me to do that, so let's start with that," or, "We'll focus on improving productivity." Some projects have broader scopes, but if they start with, "We actually want to save electricity," then their goal becomes tangible. For example, there was a company that worked with GE on an energy management project. They analyzed data coming from refrigerators and air conditioners, installed sensors in areas where data wasn't available, and analyzed the data to optimize energy savings. With a clear goal like this, they can then adopt all the relevant technologies. But if they simply say, "We need a smart factory," or "The CEO says so," they end up drifting through the review phase.

-Before introducing industrial IoT, it seems very important to first have a clear sense of purpose for the business.

A sense of purpose in the impact on your business is crucial. For example, if you spend a lot of money on maintenance, you can accurately analyze the data and cut unnecessary costs. That's why we always start with this question for companies: Let's work together to identify the desired outcome.

-So, what should companies start doing right now?

First, they need to 'GET CONNECTED'. They need to know what to do. There are quite a few companies that start this way. We believe they can gain insights we don't yet know about. One Korean company, whose name I can't reveal, thought, "We want to gather data first." They purchased an industrial database from us and stored it there for two years. This year, they began a project to analyze this data. These companies start with the belief that by gathering this much data, they can further optimize their existing areas, since they're new to the field.



-How much productivity improvement do companies experience by adopting industrial IoT solutions?

Now, about six months after implementing Predix, it's not perfect, but it's reportedly showing results. In particular, a Brazilian steel company that implemented a failure prediction solution identified two equipment locations at risk of failure. Had they failed to predict these failures, they would have incurred significant losses. Furthermore, Korean companies aren't just deploying Predix; they also have on-premise systems for data collection. GE is adopting both. Currently, the majority of its on-premise customers are still using the solution.

-How is the supply of Predix in Korea going?

The number of companies supplying Predix isn't yet in the double digits. This is because they only began in earnest last year. While last year was a cautious pilot phase, this year, more and more companies are moving towards full implementation. While all solutions built on Predix will be integrated, many still insist on on-premises deployment. In such cases, I believe customers should be given a choice.

"The experience and know-how accumulated through manufacturing is the source,
"Excellent data collection and analysis capabilities at the device level"

Besides GE, there are many companies that provide industrial IoT solutions. What makes GE different from these companies?

GE is a company with a strong presence in the manufacturing industry. Because of its extensive experience and expertise, these are embedded into its platform. First, its strength lies in its "Get Connected" strategy. While other companies' platforms focus on the cloud itself, GE recognizes the critical importance of the edge in industrial applications. Predix software components can be installed on machine sensors, controllers, and gateways. Simply install them and they collect data and send it to the cloud. Companies that started out with an IT foundation are inherently weak in this area. GE has actually created a digital twin (a virtual digital asset that corresponds to a physical asset) and integrated it into its platform. For example, when analyzing the lifespan of an industrial compressor, certain sensors must be included due to their design characteristics, and the corresponding algorithms will vary depending on the equipment. GE has developed these specifics. GE also excels in industrial analytics. Its industrial data scientists are continuously adding analytical modules to Predix. These modules aren't built from the basic platform or from statistical modules available from IT companies; rather, GE has developed a dedicated analytics module specifically for industrial use. The ability to predict when a rotating system will "die" is difficult to achieve without industry expertise.

-What is your outlook for the industrial IoT market?

Overseas global companies are eager to get started quickly, even on a small scale, and see results. Domestic companies are spending too much time focusing on perfect planning. Industrial IoT deployments require rapid execution. While planning was crucial in the past, industrial IoT has a shorter cycle and is inherently a different business. On the positive side, while last year was all about planning, this year, many companies are moving forward with implementation. I believe this year will be the first year of implementation.

GE calls itself a digital industrial company. What is its vision for the future?

The term "digital industrial company" carries significant implications. The experience itself can be applied without compromise. While we have strengths in operational technology (OT), going forward, software-based digital innovation is essential. To achieve this, GE has invested heavily and secured human resources over the past several years. GE's goal is a model that effectively integrates OT and IT.