Sensor data fusion for more reliable intrusion alarm systems
Intrusion alarm systems are currently facing a growing number of potential error sources in the environment. At the same time, alarm systems must comply with increasingly demanding legal requirements for sensors and motion detectors. As a future-proof solution, detectors equipped with Sensor Data Fusion technology raise the level of security while reducing the risk of cost- and time-intensive false alarms. This article provides a comprehensive overview of Sensor Data Fusion technology.
A cultural heritage museum in the South of Germany for decades, the installed intrusion alarm system has provided reliable protection on the premises. But suddenly, the detectors trigger false alarms every night after the museum closes. The system integrators are puzzled and conduct extensive tests of the entire system.
When they finally identify the culprit, it’s unexpected: As it turns out, the recently installed LED lighting system in the museum’s exhibition spaces radiates at a wavelength that triggers anti-masking alarms in the detectors.
Not an easy fix situation, since a new lighting system would prove far too costly. Ultimately, the integrators need to perform extensive detector firmware updates and switch to different sensor architecture to eliminate the error source. This scenario is by no means an isolated incident, but part of a growing trend.
Need for reliable detector technology
Legal requirements for anti-masking technology are becoming stringent in response to tactics by criminals
The number of potential triggers for erroneous alarms in the environment is on the rise. From the perspective of system operators and integrators, it’s a concerning development because every false alarm lowers the credibility of an intrusion alarm system. Not to mention steep costs: Every false call to the authorities comes with a price +$200 tag.
Aside from error sources in the environment, legal requirements for anti-masking technology are becoming more stringent in response to ever more resourceful tactics employed by criminals to sidestep detectors. What’s more, today’s detectors need to be fortified against service outages and provide reliable, around-the-clock operability to catch intruders in a timely and reliable fashion.
Sensor Data Fusion Technology
In light of these demands, one particular approach has emerged as a future-proof solution over the past few years: Sensor Data Fusion technology, the combination of several types of sensors within one detector – designed to cross-check and verify alarm sources via intelligent algorithms – holds the keys to minimising false alarms and responding appropriately to actual alarm events.
This generation of detectors combines passive infrared (PIR) and microwave Doppler radar capabilities with artificial intelligence (AI) to eliminate false alarm sources without sacrificing catch performance.
Motion detectors equipped with Sensor Data Fusion technology present a fail-proof solution for building security
“It’s not about packing as many sensors as possible into a detector. But it’s about including the most relevant sensors with checks and balances through an intelligent algorithm that verifies the data for a highly reliable level of security. The result is the highest-possible catch performance at the minimum risk for erroneous alarms,” said Michael Reimer, Senior Product Manager at Bosch Security Systems.
Motion detectors with sensor data fusion
Looking ahead into the future, motion detectors equipped with Sensor Data Fusion technology not only present a fail-proof solution for building security. The comprehensive data collected by these sensors also unlock value beyond security: Constant real-time information on temperature and humidity can be used by intelligent systems and devices in building automation.
Integrated into building management systems, the sensors provide efficiency improvements and lowering energy costs
Integrated into building management systems, the sensors provide the foundation for efficiency improvements and lowering energy costs in HVAC systems. Companies such as Bosch support these network synergies by constantly developing and optimising intelligent sensors.
On that note, installers must be familiar with the latest generation of sensor technology to upgrade their systems accordingly, starting with a comprehensive overview of error sources in the environment.
Prominent false alarm triggers in intrusion alarm systems
The following factors emerge as frequent triggers of false alarms in conventional detectors:
Strong temperature fluctuations can be interpreted by sensors as indicators of a person inside the building. Triggers range from floor heating sources to strong sunlight. In this context, room temperatures above 86°F (30°C) have proven particularly problematic.
Dust contamination of optical detectors lowers the detection performance while raising susceptibility to false alarms.
Draft air from air conditioning systems or open windows can trigger motion sensors, especially when curtains, plants, or signage attached to the ceilings (e.g. in grocery stores) are put in motion.
Strong light exposure directly on the sensor surface, e.g. caused by headlights from passing vehicles, floodlights, reflected or direct sunlight – all of which sensors may interpret as a flashlight from an intruder.
Extensive bandwidth frequencies in Wi-Fi routers can potentially confuse sensors. Only a few years ago, wireless routers operated on a bandwidth of around 2.7GHz while today’s devices often exceed 5GHz, thereby catching older detectors off guard.
LED lights radiating at frequencies beyond the spectrum of visible light may trigger sensors with their infrared signals.
Regarding the last two points, it’s important to note that legislation provides clear guidelines for the maximum frequency spectrum maintained by Wi-Fi routers and LED lighting.
But the influx of cheap and illegal products in both product groups – products that do not meet the guidelines – continues to pose problems when installed near conventional detectors. For this reason, Sensor Data Fusion technology provides a reliable solution by verifying alarms with data from several types of sensors within a single detector.
Beyond providing immunity from false alarm triggers, the new generation of sensors also needs to comply with the current legislature. These guidelines include the latest EN50131-grade 3, and German VdS class C standards with clear requirements regarding anti-masking technology for detecting sabotage attempts. This is exactly where Sensor Data Fusion technology provides long-term security.
Evolution of intrusion detector technology
Initially, motion detectors designed for intrusion alarm systems were merely equipped with a single type of sensor; namely passive infrared technology (PIR).
Upon their introduction, these sensors raised the overall level of building security tremendously in automated security systems. But over time, these sensors proved limited in their catch performance. As a result, manufacturers began implementing microwave Doppler radar capabilities to cover additional sources of intrusion alarms.
First step detection technology
In Bosch sensors, engineers added First Step detection to trigger instant alarms upon persons entering a room
Over the next few years, sensors were also equipped with sensors detecting visible light to catch flashlights used by burglars, as well as temperature sensors. In Bosch sensors, engineers added proprietary technologies such as First Step detection to trigger instant alarms upon persons entering a room.
But experience in the field soon proved, especially due to error sources such as rats and other animals, that comprehensive intrusion detection demands a synergetic approach: A combination of sensors aligned to cross-check one another for a proactive response to incoming signals.
At the same time, the aforementioned bandwidth expansion in Wi-Fi routers and LED lighting systems required detectors to implement the latest circuit technology to avoid serving as ‘antennas’ for undesired signals.
Sensor data fusion approach
At its very core, Sensor Data Fusion technology relies on the centralised collection of all data captured by the variety of different sensors included in a single detector. These data streams are directed to a microprocessor capable of analysing the signals in real-time via a complex algorithm. This algorithm is the key to Sensor Data Fusion.
It enables the detector to balance active sensors and adjust sensitivities as needed, to make truly intelligent decisions regarding whether or not the data indicates a valid alarm condition – and if so, trigger an alarm.
Advanced verification mechanisms
The current generation of Sensor Data Fusion detectors, for instance from Bosch, feature advanced verification mechanisms, including Microwave Noise Adaptive Processing to easily differentiate humans from false alarm sources (e.g. ceiling fans or hanging signs).
For increased reliability, signals from PIR and microwave Doppler radar are compared to determine whether an actual alarm event is taking place. Additionally, the optical chamber is sealed to prevent drafts and insects from affecting the detector, while the detector is programmed for pet and small animal immunity.
Further types of sensors embedded in current and future generations of Sensor Data Fusion detectors include MEM-sensors as well as vibration sensors and accelerometers.
Ultimately, it’s important to keep in mind that the cross-verification between sensors serves to increase false alarm immunity without sacrificing the catch performance of actual intruders. It merely serves to cover various indicators of intrusion.
Protecting UNESCO World Cultural Heritage in China
Intelligent detectors equipped with Sensor Data Fusion are protecting historic cultural artifacts in China from theft and damage. At the UNESCO-protected Terracotta Warriors Museum site, one hundred TriTech motion detectors from Bosch with PIR and microwave Doppler radar technology safeguard the invaluable treasures against intruders.
To provide comprehensive protection amid the specific demands of the museum site, the detectors have been installed on walls and ceilings to safeguard the 16,300-square-meter museum site.
To ensure an optimal visitor experience without interference from glass walls and other barriers, many detectors are mounted at a height of 4.5 meters (15 feet) above ground under the ceiling. Despite their height, the detectors provide accurate data around the clock while exceeding the performance limits of conventional motion detectors, which clock out at a mere 2 meters (6 feet) catchment area.
Integrated video systems
The site also presents additional error sources such as large amounts of dust that can contaminate the sensors, as well as visitors accidentally dropping their cameras or mobile phones next to museum exhibits.
To distinguish these events from actual criminal activity, the intrusion alarm system is integrated with the museum’s video security system. This allows for verifying alarm triggers with real-time video footage at a fast pace: In the case of an actual alarm event, the system alerts the on-site security personnel in the control room in less than two seconds.
Added value beyond security
Sensor Data Fusion technology provides a viable solution for the rising number of error sources in the environment
As of today, Sensor Data Fusion technology already provides a viable solution for the rising number of error sources in the environment while providing legally compliant building security against intruders. In light of future developments, operators can leverage significant added value from upgrading existing systems – possibly without fundamentally replacing current system architecture – to the new detector standard.
Added value how? On one hand, the detectors can integrate with access control, video security, voice alarm, and analytics for a heightened level of security. These synergetic effects are especially pronounced on end-to-end platforms like the Bosch Building Management system.
On the other hand, the data streams from intelligent detectors also supply actionable intelligence to building automation systems, for instance as the basis for efficiency improvements and lowering energy consumption in HVAC systems.
New backward-compatible detectors
Bosch will release a new series of commercial detectors by end of 2021, based on the latest research on risk factors for false alarm sources in the environment and line with current legislation and safety standards. Throughout these developments, installers can rest assured that all new detectors are fully backward compatible and work with existing networking/architecture.
With that said, Sensor Data Fusion technology emerges as the key to more secure intrusion alarm systems today and in the future.
TriTech detectors from Bosch
For reliable, fail-proof alarms the current series of TriTech detectors from Bosch relies on a combination of different sensor data streams, evaluated by an integrated algorithm. These Sensor Data Fusion detectors from Bosch combine up to five different sensors in a single unit, including:
Long-range passive infrared (PIR) sensor
Short-range PIR sensor
White light sensor
Equipped with these sensors, TriTech detectors are capable of detecting the most frequent sources of false alarms; from headlights on passing cars to a mouse passing across the room at a 4.5-meter distance to the detector. What’s more, TriTech detectors provide reliable performance at room temperatures above 86°F (30°C) while fully guarding against actual intrusion and sabotage attempts from criminals.