AI-powered alarm verification with DeepAlert software

An integration that uses AI-powered software to analyze images from Ajax MotionCam detectors for enhanced alarm verification and analysis.

About the integration

DeepAlert integrates with the Ajax ecosystem to provide additional alarm verification for MotionCam series detectors. This extra level of validation can be useful in challenging environments or when additional risks are present. While Ajax detectors effectively respond to movement, further verification can be required in cases where devices are installed without following all the recommendations, in difficult conditions, or when additional accuracy is required. 

The integration takes advantage of Ajax Cloud, allowing installers and monitoring company operators to easily connect their hubs and detectors to the DeepAlert platform via Ajax API. Cloud Signaling ensures seamless communication between Ajax systems and third-party monitoring software, converting hub notifications into event protocols compatible with the central monitoring station (CMS) software. The solution supports integration with popular third-party platforms, such as Immix and Sentinel, to provide users with the flexibility to monitor alarms using the interface they prefer.

To set up DeepAlert verification, security professionals first create an account in the DeepAlert portal and enter the Ajax hub ID. After that, they send a request to DeepAlert as a monitoring provider via Ajax apps. Once the request is received and confirmed, all detectors associated with the hub will be imported. Then, users can configure alarm delivery settings, including advanced options, such as:

• Count alert — an alarm is activated if two or more objects are detected, e.g., multiple people in the protected area.
• Scene change — an alarm is activated when camera tampering or repositioning is detected.

The number of customizable options continues to expand over time.

The process works as follows: when the device from the MotionCam series detects movement, it sends an alarm notification to the monitoring station. DeepAlert instantly analyzes the images received from the detector using AI technology and sends an additional notification confirming whether people or vehicles are in the photo. The notifications also include a probability score, such as “People detection: 95%.” In certain third-party monitoring software, it is possible to create workflows or alarm-handling processes based on the probability score. These processes can include, for example, raising alarms for the scores above 50% (high priority) and ignoring alarm notifications for the scores below 50% (low priority).

This integration is especially valuable for companies that use Ajax Ready security towers, as these devices are often deployed in varied remote outdoor environments where additional verification is important. The integration is also relevant for security operations centers (SOCs) and monitoring companies, as it allows them to accurately identify the cause of an alarm, minimize costly false dispatches, and stay focused only on real security threats.

Gallery