Page 8 - Security Today, March 2022
P. 8

Reducing Nuisance Alarms Even a low rate of alarm can lead to complacency
By Tom Hofer
Critical infrastructure sites need reliable, cost-effective physical security solutions. For these solutions to be a success they must reliably detect and deter would-be intruders, en- hance intrusion assessment and response capabilities, secure remote and unmanned sites, and, finally, address one of the most prevalent issues with physical security, re- duce nuisance alarms.
Nuisance alarms are alarms that are generated when the site’s security system is functioning correctly but is being triggered by non-threat stimuli, which may include wind or other severe weather, nearby hu- man or animal activity, or wind-induced vegetation movement.
For operators, even a relatively low nuisance alarm rate (NAR) can lead to problems such as complacency–ignoring real alarms because it is assumed they are false–and increased costs for ongoing re- pair and maintenance.
Adding a perimeter intrusion detection system (PIDS), such as a fence-mounted sensor, augments the effectiveness of an existing fence by providing valuable ca- pabilities to detect and locate attempts to cut, climb or lift the fence.
Connected to the site’s alarm systems, PIDS can notify off-site security personnel of the intrusion and engage automated de- terrent mechanisms such as the strobing of lights at the fence line, increased illumina- tion of the overall site or using public ad- dress systems to provide audible warnings.
Intrusion attempts can also be detected via video analytics by using either virtual tripwires defined along the fence or area masks. Entry-level analytics use simple mo- tion detection while more advanced analyt-
ics use sophisticated algorithms that detect and classify the presence of people, ve- hicles, and objects while rejecting environ- mental changes, including wind-induced movement, shadows and small animals.
Video analytics offer an exciting new set of capabilities that greatly enhance pe- rimeter security at relatively low cost:
• Supplement fence sensors by providing
additional detection capabilities (espe- cially important for low fences, orna- mental fences, and walls).
• Determine the direction of intrusions (ingress or egress).
• Provide covert detection (no visible equipment on fence).
• Detect and track people near both sides of the perimeter fences to provide early warning of potential security events be- fore they can occur. This early warning can be used to direct PTZ cameras, so that high-resolution video can be captured of the intruder at the time the alarm (from the fence sensor) is generated.
• Use anti-loitering analytics to detect per- sons “camped out” near the perimeter.
• Auto-track intruders with a PTZ camera af-
ter an alarm is generated (assists in assess- ment, response, and evidence collection).
Both fence-mounted sensors and video analytics are highly developed and mature technologies, but solutions that can fur- ther enhance public safety (and offer the potential to reduce operating and mainte- nance costs) warrant attention.
The concept of sensor fusion technol- ogy is not new. The basic premise of com- bining multiple sensors together to benefit from their strengths while eliminating their weaknesses has been discussed for many years. Historically, this meant a Boolean logic integration between two systems (typically a fence sensor and people track- ing video analytics, or a fence sensor and buried sensor installed in parallel).
This approach certainly can work to reduce nuisance alarms caused by well-
defined, predicted events.
A better approach is to introduce intel-
ligence into the sensor fusion system by synthesizing low-level data from the sepa- rate systems to generate actionable infor- mation. More than a simple Boolean logic integration, true sensor fusion analyzes real-time data alongside historical, loca- tional, environmental, and classification data before generating an alarm.
When signal response data is synthe- sized between sensors from fence video analytic systems, nuisance alarms generat- ed by wind or debris as well as non-threat human activity are virtually eliminated while maintaining the system’s high prob- ability of detection.
True sensor fusion has the following benefits:
• A lowest possible nuisance alarm rate leads to greater system confidence and better response times.
• Improved probability of detection (Pd), as the individual sensors can use higher sensitivity settings and announce distur- bance events faster.
• Simplified integration, as the alarm out- put arrives from a single source (sensor fusion engine).
While advances in perimeter intrusion detection and video analytic technologies will continue to offer exciting new capa- bilities, nuisance alarms will unfortunately remain an issue. Sensor fusion offers the means to defeat nuisance alarms while main- taining the highest probability of detection.
To take advantage of sensor fusion, electrical providers will need to work with security vendors that have the in-house ca- pability to intelligently synthesis low-level multi-sensor data in order to generate ac- tionable results. The resulting increase in system confidence enables security operators to focus on what’s impor-
tant, namely a quick and efficient response to real security threats.
Tom Hofer is the product manager at Senstar Corporation.

   6   7   8   9   10