On August 20, in the Tower Palace complex in Gangnam, Seoul, a quadruped robot began its patrol as intermittent rain fell. "Welcome to Tower Palace. Please be cautious as I am in motion," the robot announced repeatedly.
As it followed a designated path, the robot navigated down and up steps, avoiding pedestrians in its way and maintaining balance on inclines. The outdoor demonstration was briefly conducted under a roof due to the rain, but continued outside as the downpour eased.
The demonstration site was part of a proof of concept (PoC) for a quadruped security robot being developed by LG CNS in collaboration with Tower PMC, a premium residential management company. The two companies plan to conduct the PoC at Tower Palace for about two months before launching actual operations in the fourth quarter.
LG CNS has developed the upper equipment and patrol software, utilizing its own robot platform, 'Physical Works,' while employing the 'M20' quadruped robot from Deep Robotics for hardware. The patrol platform, 'Physical Works Patrol,' manages site maps, movement nodes, patrol routes, and photography.
The robot is equipped with both standard and thermal cameras. The thermal camera is particularly useful for detecting overheating at electric vehicle charging stations and identifying smokers. The robot travels daily along pre-designated patrol points, capturing and storing video and images, while AI analyzes the data to detect unusual situations such as intrusions or fires.
Residents passing by showed interest in the robot. One resident extended a foot in front of the robot to see if it would avoid them, and the robot recognized the obstacle, turning right to pass by. Children approached to watch its movements, and some residents took photos with their phones. One resident remarked, "This is the first time I've seen a robot move like this," expressing fascination.
Another resident tested the robot's stability by pushing it, and the robot maintained its balance on all fours. A representative on site explained that it could support the weight of a person weighing around 60 kilograms.
The robot moves at approximately 1.5 meters per second and can autonomously navigate obstacles under 15 centimeters in height.
During the PoC, LG CNS will assess the robot's safety for outdoor operations, its ability to photograph and check facility conditions, and its capacity to navigate safely in the underground parking lot, which has many steps and curbs, while detecting unusual situations. The company noted that it has reduced the robot's deployment period from several months to one or two months by enabling it to quickly learn and operate based on site data.
The technical verification in the underground parking lot will be particularly detailed, as there is a risk of large-scale accidents related to electric vehicle charging and fire hazards. The robot aims to address these issues effectively.
Using learned data, the robot will autonomously patrol common areas and the underground parking lot, alerting the management office if it detects suspicious behavior or overheating at electric vehicle charging stations. This will allow managers to respond quickly to on-site situations. The robot is also equipped with two-way voice communication, enabling real-time interaction to assess situations and determine appropriate follow-up actions.
A representative from LG CNS stated, "The robot will repeatedly patrol the electric vehicle charging area on the fifth basement level of Tower Palace, using the thermal camera to check for overheating of electric vehicles and charging facilities."
In the event of an unusual situation, the robot can automatically notify the disaster response center for prompt action. The patrol process, which typically takes security personnel 1-2 hours, can be set to repeat in shorter intervals, such as every 30 minutes.
LG CNS plans to continuously enhance the robot's functions and performance while establishing a collaborative system with various service robots. For instance, if the security robot detects trash in the complex, it will request a cleaning robot to handle the task, enabling autonomous execution of various facility management duties through inter-robot collaboration.
Myung Chang-guk, executive director of LG CNS's Smart Logistics Center/Robot Business, remarked, "This collaboration is a significant example of how physical AI technology creates customer value in real residential environments. LG CNS will continue to expand AI robot services where various robots perform actual tasks and collaborate based on Physical Works."
Kang Min-soo, CEO of Tower PMC, stated, "Among the various tasks performed by security personnel, the robot will primarily take over repetitive patrol duties. We expect to achieve significant workforce efficiency when the robot can conduct patrols reliably without human intervention."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.