다비오

▲ Park Ju Hum, CEO of Dabeeo

 

"Time equals economic cost. As time expands, so does the economic cost. While it may seem that using AI can reduce labor costs, there needs to be a conceptual shift towards establishing the idea that AI can reduce production time from a year or six months to one-tenth."

 

Park Ju Hum, CEO of Dabeeo, emphasized that the purpose of utilizing AI should not just be to reduce workforce but to increase efficiency in management by reallocating the saved time to other economic activities.

 

With the recent surge in interest and experimental applications of AI in the spatial information industry, there is a growing momentum to actively utilize AI. However, the lack of consistent regulations and policies regarding the scope of application and AI industry is leaving the industry vulnerable to the high waves of the AI era.

 

Park Ju Hum, CEO, stated, "The greatest strength of AI lies in efficiency, as it offers the opportunity to overcome the workload that humans cannot handle. However, the limitation of AI lies in the subtle differences in outcomes, where humans tend to trust decisions made by humans."

 

He added, "When making decisions, I believe there is more reliability when humans are involved than when AI is involved." He further remarked, "This is because they are experts who have been doing this work for decades, and although AI has not reached that stage yet, the breadth of trust is gradually decreasing in the current AI market."

 

The application of AI technology in land information management focuses on providing freshness by enabling extreme rapid detection of changes over large areas in a short period.

 

Park Ju Hum, CEO, stated, "The area of Seoul, 605.02㎢, can be detected within one or two hours, and even the entire national territory, approximately 100,000㎢, can be detected for changes in a short period." He added, "Once the process for AI change detection is established, the accuracy and reliability can easily exceed 90%."

 

▲ Park Ju Hum, CEO of Dabeeo

 

He remarked, "People may wonder if there could be one out of ten or twenty instances where mistakes are made in change detection, but investing in AI for 1-2 years can bring it up to that level." He foresaw, "In that case, experts will define the prerequisites for change detection, and the refinement of AI trained with those conditions will progress rapidly."

 

Expanding on this, it's explained that through the standardized and formalized processes of expert groups, reliable results can be derived.

 

Park, the CEO, also emphasized, "AI is an absolute necessity for the future," stating, "The update of land information essentially requires the premise of AI, and the entire production process should be established accordingly. Specific areas should be developed through collaboration between AI and human interpreters."

 

To enhance the reliability of AI, it's explained that a change detection AI model can be developed based on the accumulated experience and technical know-how of experts and can evolve through continuous learning.

 

Therefore, creating an ecosystem where AI and GIS experts' technical collaboration can be well integrated, allowing AI to interpret and verify data from the outset, is required.

 

Especially with spatial information being largely conducted as government contracts limited to one-year projects, which could disrupt industrial growth, overall institutional improvements are needed for sustained growth and development in the Geo AI field.

 

The viewpoint suggests that continuous investment for a certain period until achieving satisfactory results should be prioritized by the government, supporting the establishment of a production management system, and fostering a market ecosystem led by the private sector.

 

Park also pointed out, "Recently, all services of companies such as Microsoft Word, Office, and Windows are transitioning to subscription-based services," adding, "With service-based models, regardless of quality, they continue, potentially missing opportunities to utilize better technologies."

 

He explained, "If the annual license fee amounts to 1 billion won, inducing technological advancement in companies over 10 years with licenses could lead to global competitiveness enhancement, even if over 90% of countries worldwide lacking well-established maps were targeted for ODA projects."

 

Dabeeo has been in operation for 12 years as a 'Geo Intelligence Company', supporting the efficient construction of spatial information and providing applied technology for spatial planning, management, and prediction using AI technology.

 

It stands out in the fields of land information and natural environment and climate change analysis, and in the spatial information field, it specializes in indoor spaces, recently providing indoor map data for major retail spaces nationwide for KakaoMap's indoor map service.

 

Furthermore, as the only AI company participating in the Korea Land and Geospatial Information Authority's '1:1000 Digital Elevation Model Enhancement Pilot Project', it utilizes AI-based change detection technology in the Geumsan/Gunwi areas.

 

Expanding its market to global country-level data businesses, it has local branches in the United States and Vietnam.