Security company S-1 said on the 11th that it has recently introduced a series of AI and IoT-based safety management solutions tailored to the characteristics of industrial worksites.
S-1 said industrial accidents have returned to an upward trend. According to the Ministry of Employment and Labor (MOEL), 457 people died in industrial accidents from January to September this year, up 3.2% from the same period a year earlier. In particular, deaths at business sites with fewer than 50 employees increased 10.4%, and establishments with fewer than five employees surged 24.5%. The cause of industrial accidents is cited as the large number of corporations that postpone safety investments due to expense burdens. In a survey last year by the Korea Enterprises Federation, the compliance rate with obligations under the Serious Accidents Punishment Act at business sites with fewer than 50 employees was only 23%.
To prepare for this, S-1 is applying to industrial sites the following: ◇ the AI-based safety monitoring system "SVMS," ◇ the IoT sensor-based fire and gas leak detection solution "BlueScan," and ◇ the "facial recognition reader" for access control to high-risk zones.
S-1's "SVMS safety monitoring" uses an AI algorithm embedded in CCTV to analyze in real time six types of hazards—failure to wear a hard hat or gas mask, entry into restricted areas, working alone, collapse, and fire—and immediately alert the responsible staff. The Ministry of Employment and Labor (MOEL) analyzed 2,011 fatal accident cases from 2018 to 2020 and found that failure to wear protective gear accounted for more than 600 cases, making this the most basic yet easily overlooked area under manpower-based management that AI can supplement.
"BlueScan," aimed at equipment accidents such as fires and gas leaks, is also spreading mainly among small and midsize manufacturers. IoT sensors are attached to key equipment such as generators, electrical rooms, and machine rooms to immediately detect water leaks, power outages, and gas leaks, and automatically notify the responsible staff and fire authorities when an abnormality occurs. It is cited as advantageous for significantly improving management efficiency relative to expense at business sites where staffing at night and on holidays is difficult.
The "facial recognition reader" for access control to high-risk zones is also being adopted on manufacturing floors. With Deep Learning-based facial recognition technology, it identifies individuals with 99.97% accuracy and blocks unauthorized entry into restricted areas. Given that workers wearing safety gear often find fingerprint or card tagging difficult on site, facial recognition is evaluated as providing both security and convenience. Authentication takes just 0.6 seconds, and up to 50,000 people can be registered, making it applicable to large industrial complexes.
An S-1 official said, "As industrial safety has emerged as a social issue, a preventive system based on AI technology is receiving more attention than ever," and added, "Based on 48 years of know-how protecting key national facilities and AI technology, we will help create a safe industrial environment where corporations can focus on business activities with peace of mind."