Improving Security for Problematic Zones
When helping customers secure their sites via the use of perimeter intrusion detection technologies (PIDS), Senstar is often asked about how to better handle problematic zones, namely segments of the perimeter fence that threat assessments identify as being at a greater risk of attack or areas that traditionally have generated unacceptably high nuisance alarm rates (NAR).
What Makes a Zone High Risk?
High risk zones are segments of the perimeter fence that an intruder may view as being easier to break into than others. For example, nearby heavy vegetation or poor lighting may provide partial concealment, while segments further away from buildings or high traffic areas may offer a sense of lower monitoring. Risk may also be introduced by limitations with the existing security implementation, such as changes in fence construction or by distractions caused by regular business operations, such has having authorized activity at a gate provide cover for a nearby intrusion.
To mitigate these risks, the sensitivity of sensors or analytics may be increased, or the alarm declaration threshold lowered, in order to ensure detection in these edge cases. However, with traditional intrusion detection technologies, the additional detection capabilities may result in an increase in nuisance alarms.
What Generates Nuisance Alarms?
False or nuisance alarms are an on-going concern for customers (and the security industry in general). High nuisance alarm rates can lead to distractions, information overload for monitoring staff, costly physical visits to the perimeter, and eventual mistrust in the site’s security systems.
Intrusion detection technology from a respected vendor, be it a fence sensor, buried sensor, microwave, or video analytic, should not normally generate excessive nuisance alarms when installed in accordance with manufacturer specifications (Senstar’s fence-mounted sensor, FlexZone, for example, has proven abilities to avoid nuisance alarms even under extreme conditions). However, there are situations where nuisance alarms may still be generated and are difficult to resolve via standard security technologies. For example, depending on its construction, a swinging gate may generate “noise” under specific wind conditions, causing problems with fence-mounted sensors, while nearby authorized activity, vehicle headlights, shadows, or vegetation movement may generate unnecessary video analytic events.
To address these concerns, the common approach is to adjust sensor settings to identify specific characteristics such as noise, or to limit the range of data captured to avoid non-threat input (for example, masking out all but the most critical areas of a video scene to avoid false positives). Another approach is to simply combine multiple sensors together in a Boolean logic AND configuration, so that two separate systems must both agree before an alarm is generated.
These approaches can work, but introduce potentially negative consequences:
- Adjusting detection settings to avoid nuisance alarms can potentially lead to missed detections against for sophisticated attacks or specific edge cases (note that to address these edge cases, additional security mechanisms can be deployed, such as enhanced fence-toppers and double-pass sensor runs).
- Factors outside the security department’s control (changes in vegetation, weather, new construction, or even daily schedules) can introduce new sources of nuisance alarms or missed detections, requiring frequent adjustments or re-calibrations.
- The logical ANDing of multiple sensors adds additional integration complexity, raises deployment concerns, but mostly importantly, can result in critical missed detections under specific edge cases (worst outcome!).
Sensor Fusion: Solving Real-World Problems
Senstar, with its 40+ years of experience bringing innovative technologies to market, has developed sensor fusion technology as a novel approach to dealing with nuisance alarms, with the goal of defeating them once and for all. Not a simple Boolean logic combination of sensor outputs, Senstar’s sensor fusion engine intelligently synthesizes low-level data from fence sensors and video analytics to characterize risk and make context-aware decisions.
A key use case for sensor fusion is resolving security issues for problematic zones. Consider a composite site with the following characteristics (based on real issues our customers have faced):
- The perimeter is protected by a security fence equipped with a fence sensor. While the overall system works well, a swinging gate at the entrance to the site moves or shakes when exposed to frequent high-wind conditions, generating nuisance alarms.
- Behind the site is a heavily forested area. Intruders can use this forest to conceal their approach, making that section of the perimeter a prime target for intruders. To combat this threat, an increased level of security is required.
- Remote monitoring staff maintain situational awareness via a common operating platform that unifies video surveillance, sensor alarms, video analytics, and access control.
To better protect the two problematic zones, sensor fusion can be enabled specifically for just them, leveraging the existing investments in fence sensors and video surveillance and avoiding the costs and complexities associated with integrating multiple detection systems together. This approach, adding sensor fusion capabilities to only the zones that require it, avoids complexity, simplifies operations, and manages costs.
The result is improved security for the forested zone, and a virtual elimination of nuisance alarms from the swinging gate zone. For the remoting monitoring staff, the new sensor fusion functionally is completely transparent, with them only receiving one alarm per valid intrusion attempt and helps to ensure any generated alarm is in response to a valid security threat and requires immediate attention.