Live demonstration of a large commercial facility

Connected to a large commercial facility in Australia, visitors were able to see ``icetana'' in operation in real time.

The participants understood what kind of events are judged as abnormalities in actual commercial facilities, and how wide a range of abnormalities can be detected.

When we explained the features of icetana, the one that received the most surprising and convincing comments was its ability to display only images that icetana has determined to be abnormal on the monitor. Monitors at the disaster prevention center no longer need to view a huge number of camera images, and only check for ``abnormalities'' displayed on the monitor. This feature allows security guards to react without overlooking any abnormalities, and to take appropriate measures immediately afterwards. Security company visitors also say, ``I see, this is... I received a comment saying, ``That's interesting.''

We also introduced the highlight function. We showed that it is possible to filter and search for past abnormalities, and efficiently detect a wide range of abnormalities such as falls, running the wrong way on escalators, staying in the car, and dangerous acts. We also introduced the operational feel of the UI and gave the participants a hands-on experience of its use.

Mitsubishi Estate implementation example

We also introduced a use case of Mitsubishi Estate, which is actually using "icetana" in Japan, in a video. It is currently in operation at the Maru Building and Shin-Maru Building near Tokyo Station, and was adopted as one of several technologies introduced by Mitsubishi Estate in pursuit of a "next-generation facility management model."

What kind of social issues are they approaching? What kind of social issues are they approaching? How are demonstration experiments being conducted before the introduction? What are the voices of security guards? Also, are there events that are difficult to set up using rule-based anomaly detection (e.g., a crow attacking a person). ), etc. Please click here to view the video and the full text of the white paper.

Mitsubishi Estate's challenges supported by next-generation anomaly detection software "icetana"