"ONE FUKUOKA BLDG." (commonly known as "One Bldg.") was born out of "Tenjin Big Bang," the redevelopment project in Tenjin, Fukuoka. It is a large-scale mixed-use complex of 19 floors above ground that integrates retail, offices, a hotel, and a conference hall, and it is visited by many people every day.
Nishitetsu Building Management Co., Ltd., which is responsible for managing the facility, introduced the AI security system "AI Security asilla" in time for the facility's opening in April 2025 and has built a security framework that leverages AI.
For this article, we spoke with Mr. Masayuki Kamezaki, Manager of the Security and Crime Prevention Department at Nishitetsu Building Management Co., Ltd., about the background to introducing AI security, the effects they are experiencing on-site, and their outlook for the future.
.png)
.png)
Kamezaki: ONE FUKUOKA BLDG. is a facility born as the largest-scale redevelopment project within Tenjin Big Bang. It has 19 floors above ground and 4 floors below, with a total floor area of approximately 147,000 square meters. It attracted a great deal of attention even before opening, and since opening as well, we have had a truly large number of visitors every day — including people who come for tenants opening their first stores in Kyushu, as well as inbound visitors.
With this scale and number of visitors, the security front line constantly needs to prepare for a wide range of risks. There are about 500 cameras within the facility, but there is a limit to monitoring that many with human eyes alone. Even though we had the cameras, we weren't able to make full use of them. To be honest, it was hard to say that we had a sufficient framework for actually "watching" the cameras.
A major theme was how to raise security quality with a limited framework. In particular, during the business hours of the retail areas there are many visitors, and the burden on dedicated patrol staff was significant.
Kamezaki: With the Nishitetsu Group as a whole promoting DX and ICT, we felt that this wave had now reached the security field as well. Precisely because it was a new facility, we on-site had expectations that we might be able to carry out security that leverages the latest technology.
AI Security asilla is a system that utilizes the cameras we already use as-is, with AI analyzing the footage and notifying us if there is an anomaly. Because there was no need to replace cameras or carry out large-scale construction, we were grateful that we could introduce it without stopping on-site operations. Within a limited framework, AI joins us as "another set of eyes." For the security staff on-site, it's a very reassuring presence.
Kamezaki: Yes. In about three months we were able to use the system well and incorporate it into daily operations. The fact that our on-site members are relatively young, and therefore had little resistance to a new system, was also a big factor.
Asilla's support framework is also attentive and their responses to questions are quick, which helps. In addition, we can report to the owner persuasively based on operational data and detection results, which is very reassuring.
Kamezaki: Looking back over this past year, I feel that things would have been completely different with versus without AI.
The effect I feel most is that the initial response to anomalies has become dramatically faster. When an anomaly such as a fight, an act of violence, or a fall occurs, AI Security asilla instantly sends an alert from the security camera footage. The alert displays the footage at the time of detection alongside real-time footage, so even while at the disaster-prevention center we can check "what is happening on-site right now" and immediately judge whether a response is needed.
Moreover, AI monitors 24 hours a day, 365 days a year without rest, so even while we are on patrol or doing administrative work, it properly notices events that human eyes might overlook. The fact that the psychological burden of "we might be missing something" has been greatly reduced is a fairly big change for the front line.
Kamezaki: We use it largely for the purposes of "safety management" and "responding to nuisance behavior."
First, on the safety-management side, we use AI mainly around escalators and parking areas, which are especially high-risk within the facility. In fact, there was a case where we detected a fall on an escalator and were able to respond quickly. AI also detected cases such as a foreign visitor passing through a gap in a partition and entering the wrong flow line.
It's significant that AI picks up events that might otherwise have been overlooked without anyone noticing with human eyes alone.
Kamezaki: Partly because of our location in the Tenjin area, there are people who ride skateboards in prohibited areas within the facility. This can lead to the risk of injury from contact with pedestrians, as well as damage to the facility. By responding promptly each time AI detects it, the number of people skateboarding within the facility has gradually decreased. As a result, this has also led to a deterrent effect.
AI also detects and notifies us of intrusion into areas outside business hours. Being able to grasp and respond to such nuisance behavior immediately, without relying on the timing of patrols, is reassuring for the front line.
Kamezaki: At night we conduct security with a framework that has fewer people compared to daytime. For that reason, the presence of AI is very significant. AI firmly covers parts that are hard to keep an eye on with few staff — such as detecting intrusion into areas outside business hours and detecting anomalies in the exterior grounds. I feel that the level of nighttime safety management has definitely risen thanks to AI.
Kamezaki: Over this past year, we have fully confirmed the effectiveness of AI Security asilla. Going forward, we want to further increase the number of target cameras and eliminate blind spots within the facility.
One new attempt we are considering is advancing security through integration with speakers. Because the facility is large, in areas at a distance from the disaster-prevention center, even if we notice an event, it inevitably takes time for security staff to rush over. AI Security asilla has a speaker-integration function, so if we could automatically play a warning announcement at the moment a target event is detected, we believe we could make the initial response even faster.
To be honest, the labor shortage is advancing across the entire security industry, and we don't know whether the current framework can be maintained indefinitely. In that context, we leave to AI what AI can do, and we concentrate on the judgments and responses that only people can make. I believe that firmly building such a framework will lead to continuing to protect the safety and security of ONE FUKUOKA BLDG.
"AI Security asilla" Goes into Operation at "ONE FUKUOKA BLDG.," a New Landmark in Tenjin, Fukuoka
https://jp.asilla.com/post/one-fukuoka-building-20250610
.png)
.png)
.jpg)