Sep 29, 2023
In today's interconnected and digital world, safeguarding critical infrastructure (CI) is
of paramount importance. Facilities such as airports, utilities, rapid transit systems, and seaports play a pivotal role in our society, and they provide the backbone of commerce and our economy. Threats such as crime, terrorism, weather, natural disasters, drones, and weak points in the perimeter pose risks that need to be addressed.
Without a well-functioning physical security system, CI facilities can be compromised and rendered inoperable. Security breaches can result in significant damage, and economic losses. Potential impacts include power outages, data loss, operational disruptions, damage to reputation, and compliance violations. All of these consequences reinforce the importance of revolutionary, robust perimeter security.
Traditional perimeter surveillance systems have long relied on human vigilance to monitor cameras and process an unmanageable quantity of nuisance alarms. To truly protect a perimeter takes extraordinary human resources and even then, protection is marginal at best. They are typically missing an array of automation and AI capability for such a serious portion of security!
Enter Automation and AI-Boosted Video Analytics
By leveraging the capabilities of automation, artificial intelligence, and advanced video analytics, solutions are now capable of distinguishing between normal environmental motion and benign activities and potential security breaches. AI video analytics reinforces perimeter protection by preventing or detecting illicit entry thereby averting potential disruptions to essential services.
What sets automated AI-boosted video analytics apart is its ability to adapt and learn from real-world scenarios. Environmental and Machine learning algorithms enable these systems to continuously improve their accuracy in detecting threats while reducing nuisance alarms. This dynamic self-improvement mechanism ensures that security personnel can trust the alerts they receive, focusing their attention on genuine threats.
Nuisance alarms are a specific type of false alarm that occur when the system triggers an alert or alarm in response to an event that is not actually a threat, but still falls within the parameters of the system's detection criteria. These types of alarms may, or may not, require action by the security team. Examples include wildlife that breaches the perimeter, moving vegetation, headlights from vehicles, or shadows cast from clouds.
With modern automated and AI-boosted video analytics, most existing monitoring systems (VMS, PSIM, etc.,) can be enhanced to accurately detect, track, classify, present, and deter intrusions. The combination transforms ordinary security cameras and other sensors into autonomous, accurate, and scalable perimeter intrusion detection systems, delivering real-time situational awareness. More advanced video analytics systems can provide the geolocation of the intruder on GIS map systems for increased situational understanding. Further functions can be enabled through integration with numerous products, including but not limited to other VMS/PSIM software, ground-based RADARs, microwaves, fence, and intrusion sensors, access control, building management, and radio and other communication systems. Finally, standard management features such as policies for alarms, automated lock-down actions, self-diagnostic health monitoring, and automatic alarm routings can all be implemented.
Not All Video Analytics Are the Same
Processing nuisance alarms is not only timely, but also expensive. For an operator to effectively perform their job, they need a highly accurate system that will automatically distinguish between real threats and nuisance alarms, or auto-verify.
At PureTech Systems, we have developed patented, automated AI-boosted geospatial video analytics algorithms and employed a range of techniques to minimize nuisance alarms. By continuously improving and testing our algorithms we provide our customers with the highest level of accuracy and reliability in perimeter protection.
These algorithms are designed to automatically distinguish between real threats and normal activity. PureTech’s Auto-Verify uses Deep Learning Neural Nets to classify the detected object. If the detected objects classify as an object of interest, then and only then, is an alarm issued. Our algorithms leverage a range of factors, such as the size, shape, geo-location, and movement patterns of objects in the video feed, to determine whether an event is a real threat without the need for operator intervention. When integrated into a new or existing enterprise surveillance system, it allows for security personnel to effectively monitor their critical infrastructure facility.
The advancement of automations and AI-boosted geospatial video analytics has ushered in a new era of critical infrastructure protection. An unparalleled level of detection accuracy and the elimination of nuisance alarms is vital in transforming the way we secure airports, utilities, rapid transit systems, and seaports.
PureTech will be at the following shows in 2023: GSX, Intersec and DSEI.