Artificial intelligence (AI) is changing every industry. AI is being applied in almost all areas, including Korea's representative core industry, semiconductors, as well as entertainment, beauty, and healthcare. There are also forecasts that artificial general intelligence (AGI), which possesses human-level intelligence, will become a reality within a few years.
ChosunBiz noted that changes incorporating AI are being made even in the construction industry, one of the most traditional sectors. Accordingly, a special report titled 'AI and the Construction Industry' has been prepared to examine the changes that AI is bringing to various aspects of the construction industry that creates our living spaces. To explore how AI is being integrated on construction sites, experts' opinions will also be shared.
As the camera illuminated the cracks on the road, the artificial intelligence (AI) program identified the cracks and marked them in red. It didn't stop at identifying the cracks but also analyzed and conveyed detailed data, such as changes in the cracks compared to the transfer.
When images of the upper part of the soil retaining structure, which are difficult to check with the naked eye, were taken by a drone, it also played a role in interpreting whether the anchors were accurately embedded.
At the Lotte Engineering Technology Research Institute in Nonhyeon-dong, Gangnam-gu, Seoul, visited on the 23rd, it was possible to see the AI technology that Lotte Engineering is implementing through visual materials. Lotte Engineering has been continuing relevant research since 2023 by creating a dedicated AI organization composed of R&D personnel and business unit staff.
◇The industry's first AI-based soil retaining structure stability assurance
Lotte Engineering developed a 'soil retaining structure rear crack tracking system' that can visualize cracks occurring in the rear of AI-based soil retaining structures (nearby buildings, roads, etc.) since 2023 and filed a patent for related technology.
Soil retaining structures are facilities installed to prevent earth collapse or groundwater outflow when excavating underground at construction sites. According to the Ministry of Land, Infrastructure and Transport, out of 63,149 risk incidents in construction sites last year, 23,971 incidents, or 38%, occurred in temporary structures.
To prevent risks occurring in temporary structures, Lotte Engineering became the first in the industry to utilize AI-based video analysis technology. When a construction worker records site videos with personal recording devices (action cameras, mobile phone cameras, etc.) and registers them on the platform, the AI model visualizes video analysis and crack information, allowing for comparative analysis of crack progression over time through history management.
This significantly reduced the manpower and time required for crack detection. Additionally, accuracy improved greatly compared to visual assessments.
In fact, cracks that appeared indistinguishable to the reporter visually were analyzed in detail by AI, which provided numerical data. In the case of the anchors, it was able to determine their accurate positioning even from just images.
Lotte Engineering has partnered with 'VisionIn', a company developing AI-based intelligent video analysis solutions, to enhance AI training speeds. To shorten the time required for AI learning from videos and images taken on site, VisionIn provides virtual images for training. Once the AI training process is complete and a certain level of accuracy is secured, there are plans for nationwide application at all sites.
Not only can the crack reading of the soil retaining structure be performed, but design of the structures can also be done through AI. By marking the construction site range on a map, AI analyzes factors such as slope to proceed with the design. A representative from Lotte Engineering stated, 'In the past, design changes occurred frequently, and with limited resources and personnel, it was time-consuming to individually review the safety of construction sites across the country,' adding, 'Utilizing AI has allowed us to reduce time and select the optimal cross-section of temporary structure designs with fewer personnel.'
◇Establishment of safety situation centers for nationwide site monitoring from headquarters
In 2023, Lotte Engineering opened a comprehensive video control system called 'Safety Situation Center' linked to an AI system at its headquarters. By utilizing the 'Risk Assessment AI System' developed with Lotte Information & Communication, it selects high-difficulty sites based on analyzed data and monitors them closely. During weather warnings such as typhoons and earthquakes, it also checks the site response status to prepare for emergencies.
CCTV installed at all Lotte Engineering sites is monitored in real-time from headquarters to detect incidents and prevent accidents, with safety management experts minimizing blind spots through recordings taken from various angles.
Last year, work began on developing AI safety monitoring technology based on worker behavior to prevent safety accidents. Ahead of technology application, a demonstration held in December included workers wearing a total of six wearable sensors, performing tasks in scenarios such as heavy equipment collisions, entrapments, abnormal worker behavior patterns, and health monitoring during commuting. Continuous monitoring data is secured through fixed cameras, which is analyzed using an AI safety management platform equipped with AI models, aiming to prevent on-site safety accidents and reduce incidents.
◇Enhanced corporate security through the development of a proprietary chatbot
As issues surrounding the leakage of corporate secrets and personal data through commercial chatbots like ChatGPT have surfaced, corporations have begun developing their own AI to prevent this. Lotte Engineering has also developed its own chatbot 'Q-bot', which has been in use since the 2nd of this month. Q-bot currently offers four services: ▲Q-Quality Case Bot ▲Q-Quality Case (Plant) Bot ▲Q-Quality Manual Bot ▲High-rise Tower (Chimney Case) Bot.
When a work-related question is entered into Q-bot, it analyzes vast amounts of data such as work guides and construction quality cases to provide relevant information. Recently, concerns about data leakage arose when new employees at corporations began searching work guides using commercial chatbots; however, Q-bot mitigates this risk by operating on a server managed independently by Lotte Engineering.
When unrelated questions related to construction work were asked during a program demonstration on site, the response indicated that there was no related information available. When requested to simply explain various construction terms, about three lines of summarized information were provided.
A representative from Lotte Engineering stated, 'We are currently operating four services on a trial basis, and employees are showing positive reactions,' adding, 'Through further development, more services are expected to be offered by 2026.'