Güvenlik Online Logo
Technical Article

AI for Illegal Dumping Monitoring: Human Review After the Alert

Sistem Yöneticisi4 min read

The APAMAN trial described by Brivo shows how an illegal-dumping alert can lead to action through staff review and a request to remove the waste. However, the case accounts do not measure the system’s detection performance on their own.

The APAMAN trial described by Brivo shows how an illegal-dumping alert can lead to action through staff review and a request to remove the waste. However, the case accounts do not measure the system’s detection performance on their own. 

The Alert Does Not Make the Decision; the Manager Reviewing the Footage Does

At a property where waste has been dumped illegally, AI’s role is not to make the intervention decision automatically, but to direct the manager to footage that needs review. In the cloud-based system described by Brivo, an alert is sent when footage matching a defined event is detected; staff review the recording on a computer or phone and decide what to do. For a property manager, this is the key workflow to assess: detection, verification and then intervention.

The problem this approach targets is less the presence of waste than the fact that the incident is noticed too late. According to the source, managers cannot investigate every case as it happens; by the time an item is found, it may be unclear who left it and when. From an editorial perspective, the real value is therefore not merely a recording showing the waste, but access to footage that can help establish how the incident unfolded. [1]

Natural-Language Descriptions, Remote Checks

Brivo says that situations to be detected can be described in natural language. The system analyzes camera footage against structured detection criteria and generates a notification for a matching event. Here, natural-language description is a way for managers to tell the system what they are looking for; the operation described by the source still relies on predefined conditions. Reviewing footage after a notification remains the responsibility of staff.

For teams managing multiple buildings, one possible use case is checking the alert’s contents before going to the site. The manufacturer says Brivo Eagle Eye Video supports compatible existing cameras and can centrally manage footage from different properties. This does not mean that every existing camera can be used: teams must separately verify which cameras are supported for their project. Nor should remote review be interpreted as eliminating site visits in every case. [1]

The Trial’s Results Must Be Kept Separate from Other Cases

According to Brivo’s account, Presto Service, a partner of the APAMAN group, is trialing the system in residential properties it manages. In one incident described as part of the trial, an AI agent named Eeva generates an alert; staff review the footage and confirm that bulky waste was left in the rubbish collection area. A removal request is then posted, and the waste is said to have been removed three days later. The outcome in this example is described as a process dependent not only on detection, but also on staff verification and action following the notice.

The source also reports, in separate operational examples from Fukuoka, that items left after a warning notice were removed by those responsible, and that vehicle footage in some cases helped with reporting to police and identifying those responsible. However, the source note distinguishes these examples from the trial APAMAN announced on 15 October 2026. It would not be accurate to treat them all as the combined success of the same pilot. [1]

The Limits of Case Studies in Purchasing Decisions

For managers, the takeaway should be to treat these examples as a draft workflow, not as a measure of detection accuracy or savings. The material provided offers no comparative accuracy, false-alarm or cost measurements. The manufacturer also notes that results may vary depending on the property and operating conditions. Any potential trial should therefore establish in advance how staff will verify alerts and what intervention each alert will trigger; individual outcomes described here should not be treated as guarantees for another building. [1]

Sources

Source: Brivo

More on the topics covered in this piece

Cookie preferences

Essential storage supports site features and remembers your choice. Consent-based Google analytics depends on your choice and separate activation. Advanced Mode is off by default; if separately enabled, it may send cookieless signals under denied consent. If activated, Umami measures pageviews independently of cookie choices. Read the Cookie Policy

Essential — always active

Session security and remembering your cookie preference.

Usage measurement for pages, searches and sponsor interactions; Google Analytics / Google Tag Manager when configured.

Use of advertising-related cookies. Contextual sponsor placements may still appear without this permission.

Preference for sharing user data with Google advertising services. No such transfer occurs unless separately enabled.

Preference for Google ad personalization, independent of analytics permission.

You can change or withdraw your choices using Cookie preferences in the footer.