Shin Kong Security Expands Smart City Initiatives Targeting AI Behavior Recognition Opportunities

The 2026 Smart City Exhibition is approaching, and Shin Kong Security (9925) has unveiled several innovative AI behavior recognition applications at a pre-exhibition press conference on March 10. The company is targeting high-demand sectors such as financial fraud prevention, retail theft prevention, home security management, and medical logistics, showcasing the transition of the security industry from “passive monitoring” to “active alerting”, actively seeking opportunities for upgrading safety in smart cities.

To implement the policies set by the Financial Supervisory Commission aimed at strengthening ATM fraud prevention and blocking fraudulent cash flows, Shin Kong Security has recognized the challenges faced by customers, such as fraudsters deliberately obscuring their facial features to evade detection. To combat this, they have introduced the “AI Facial Recognition Access Control System”. Before accessing the automatic service area of a branch ATM, users are required to remove masks, helmets, or sunglasses. The system will only unlock the access once it confirms that the user’s face is fully visible. Additionally, by integrating video feeds with access control, the system significantly enhances retrospective investigation capabilities.

For open ATM environments, Shin Kong Security has launched the “ATM Fraud Prevention AI Recognition System”. When users insert their card and enter their PIN, the ATM automatically activates AI image recognition. If the system detects that the user’s face is obscured, it will provide an automated voice prompt, saying, “Please remove your hat, mask, or sunglasses to reveal your full face.” The system can also pause transactions, in line with financial institutions’ requirements. This proactive alerting technology effectively strengthens fraud prevention measures within financial environments and captures market opportunities for smart recognition solutions.

In retail and office environments, Shin Kong Security has introduced the “Spaces AI Risk Alert Solution.” This solution utilizes Vision-Language Models (VLM) technology, allowing surveillance systems to evolve from simple object recognition to “semantic understanding.” The AI can not only recognize labels but also comprehend complex behaviors. For instance, if a scenario arises where a “product is stuffed into a coat,” “goods are taken without payment,” or “the cash register is being tampered with,” the system can proactively assess the situation and promptly alert staff, enabling immediate intervention before financial losses occur.

Shin Kong Security emphasizes that continuous AI behavior analysis can effectively fill the gaps in manual inspections, helping businesses shift from post-event video review to proactive prevention. The solution offers flexible customization for various sectors, including chain retail, logistics warehousing, and restaurants, assisting businesses in achieving digital transformation and operational resilience in a labor-short environment.

According to statistics from the Ministry of Health and Welfare, falls are the second leading cause of accidental injury deaths among seniors aged 65 and above, making it a critical issue in an aging society. Shin Kong Security has launched the “Event-Based Vision Sensor (EVS) Fall Detection System.” Utilizing edge AI technology, it analyzes dynamic contours in real-time. Once a fall is detected, the system immediately triggers an alarm mechanism to notify management or family members, ensuring timely intervention during critical moments.

The company points out that this system operates “without cameras” and does not capture images, thus alleviating privacy concerns. It can also overcome challenges posed by humidity and lighting variations, making it suitable for private spaces such as bathrooms and bedrooms. Beyond long-term care facilities and wellness villages, the system can also be adapted for social housing and regular households, serving as an invisible safety net for seniors living alone. With the healthcare market for the elderly continuing to expand, smart care solutions are expected to become a significant growth driver in the future.

In terms of automation applications, Shin Kong Security showcased two models of Autonomous Mobile Robots (AMRs) at the press conference. By integrating navigation and path planning technologies, these robots can perform tasks such as greeting, guiding, and delivering supplies in indoor environments like restaurants, medical facilities, and offices, thereby reducing labor burdens and enhancing operational efficiency. This highlights the diverse potential applications of AI in intelligent environments. As service industries and healthcare systems face long-term labor shortages, the demand for autonomous mobile robots continues to increase, becoming one of the key solutions for upgrading smart environments.

One of the AMRs is equipped with “access control design”, which combines multiple permission controls and medication order verification mechanisms to execute controlled drug delivery processes in healthcare facilities. This adheres to the medical operational principles of “three checks and five confirmations,” effectively relieving nursing staff of the burden of in-hospital medication delivery. Shin Kong Security states that the AMR robots can be flexibly deployed according to the needs of different environments, with applications extending to hospitality, office buildings, social housing, shopping malls, and public transport stations, providing a range of smart services.

Shin Kong Security emphasizes that by incorporating AI behavior recognition and image analysis technologies, security systems have evolved from passive video review to real-time judgment and proactive alerting mechanisms. The company plans to continue enhancing its AI behavior recognition and system integration capabilities and will unveil new products and branding videos during the Smart City Exhibition to showcase the results of its intelligent security initiatives.