With the ever-growing availability of video data thanks to the low cost of high-resolution video cameras and storage, artificial intelligence (AI) and deep learning analytics now have become a necessity for the physical security industry, including access control and intrusion detection. Minimizing human error and false positives are the key motivations for applying AI technologies in the security industry.

What Is Artificial Intelligence?

Artificial intelligence is the ability of machines to learn from experience using a multi-layer neural network, which mimics the human brain, in order to recognize items and patterns and make decisions without human interference.

The human brain is estimated to have 86 billion neurons; in comparison, the newest Nvidia GPU Volta has 21 billion transistors (the equivalence of a neuron), which offers the performance of hundreds of CPUs for deep learning.AI can learn continuously 24 hours per day every day, constantly acquiring, retaining and improving its knowledge

In addition, unlike humans, AI can learn continuously 24 hours per day every day, constantly acquiring, retaining and improving its knowledge. With such enormous processing power, machines using Nvidia GPU and similar chips can now distinguish faces, animals, vehicles, languages, parts of speech, etc.

Depending on the required complexity, level of details, acceptable error margin, and learning data quality, AI can learn new objects within as fast as a few seconds using Spiking Neural Network (SNN) to a few weeks using Convolution Neural Network (CNN). While both SNN and CNN offer advantages and drawbacks, they outperform tradition security systems without AI in terms of efficiency and accuracy.

According to the research reports of MarketsandMarkets, the market size of perimeter intrusion detection systems is projected to increase from 4.12 billion USD in 2016 to 5.82 billion USD in 2021 at a Compound Annual Growth Rate (CAGR) of 7.1%.

Meanwhile, the predicted market of AI in security (both cyber security and physical security) will grow from 3.92 billion USD in 2017 to 34.81 billion USD by 2025, i.e., with an impressive CAGR of 31.38%.

Legacy Perimeter Intrusion Detection Systems

Legacy perimeter intrusion detection systems (PIDSs) are typically set up with the following considerations:

  • Geographical conditions: landscape, flora, fauna, climate (sunrise, sunset, weather conditions, etc.), whether there are undulations in the terrain that would block the field of view of cameras
  • Presence or lack of other layers of physical protection or barriers
  • Integration with other systems in the security network: camera, storage, other defensive lines (door, lock, alarm, etc.)
  • Types of alarm triggers and responses
  • System complexity: intrusion detection with various types of sensors, e.g., microwave sensors, radar sensors, vibration sensors, acoustic sensors, etc.
  • Length of deployment
  • Local regulations: privacy protection, whether the cameras/sensors must be visible/hidden/buried, etc., electromagnetic interferences that may affect other systems such as oil rigs or power plants
  • Human involvement: on-site personnel arrangement, human monitoring, human action in response to alarms

The amount of false alarms can be reduced by 70% to orders of magnitude with AI
AI object detection can easily distinguish different types of people and objects

Pain Points In Intrusion Detection Systems And Benefits Of AI

The conditions listed above correspond to certain requirements of an intrusion detection systems: minimal false alarm, easy setup and maintenance, easy integration, and stable performance.AI by nature is designed to learn, adapt itself and evolve to work in multiple conditions: it should be integrated with existing video recording systems 

Minimal false alarms: False alarms lead to increased cost and inefficiency but are the main problem of PIDSs without AI technology, where animals, trees, shadows, and weather conditions may trigger the sensors.
AI object detection can easily distinguish different types of people and objects, e.g., in a region set up to detect people, a car driving by, a cat walking by, or a person’s shadow will not trigger the alarm. Therefore, the amount of false alarms can be reduced by 70% to orders of magnitude.

Easy setup and maintenance: Legacy PIDSs without AI must account for terrain, line of sight of cameras, sensor locations; any changes to the system would require manual effort to recalculate such factors and may disturb other components in the system.
In contrast, AI PIDSs enable the system administrator to access the entire system or individual cameras from the control room, configure the region and object of interest in the field of view of cameras within minutes, and adjust with ease as often as necessary. Computing knowledge and even specific security training are not required to set up a secured PIDS with AI because AI PIDS is designed to relieve humans from knowing the inner working of machines.

Easy integration with complementary technologies: Legacy PIDS without AI relies on physical technology, which are often proprietary and require complete overhaul of systems to function smoothly. On the other hand, AI by nature is designed to learn, adapt itself and evolve to work in multiple conditions, so AI PIDS is easily integrated with existing video recording (camera) and storage (NVR) systems.
AI also eliminates the need for physical wireless or fiber-based sensors; instead, it functions based on the videos captured by cameras. Furthermore, AI enables easy and instantaneous combinations of multiple layers of defense, e.g., automatic triggering of door lock, camera movement focusing and access control as soon as a specified object is detected in the region of interest, all set up with a click of a button. 

Stable performance and durability: Legacy PIDS without AI requires complicated setup with multiple components in order to increase detection accuracy. More components mean a higher probability of malfunction in the system, including exposure to damages (e.g., sensors can be destroyed) and delay in detection, while human monitoring is inconsistent due to human fatigue (studies have shown that a person can concentrate in mundane tasks for only up to 20 minutes, and the attention span decreases even more rapidly when humans are faced with multiple items at once, e.g., multiple camera monitoring screens).
AI significantly reduces, if not completely eliminates the need for human involvement in the intrusion detection system once it is set up. In addition, AI reduces the risk of system malfunction by simplifying the hardware sensors needed.   

The most advanced AI intrusion detection system today provides an all-in-one solution to distinguish any combination of alarm-triggering criteria
Minimizing human error and false positives are the key motivations for applying AI technologies in the security industry

Additional Benefits Of AI In Intrusion Detection

Artificial Intelligence is undeniably reshaping every business and weaving into every aspect of daily life Maximal detection capability: The most advanced AI intrusion detection system today provides an all-in-one solution to distinguish any combination of alarm-triggering criteria beyond perimeter protection. Using AI, the system administrator can configure as many zones with different settings and object of interests as necessary, which include detections for specific colors or attributes (e.g., person not wearing the required uniform or carrying food/drink), numbers and dwell time (e.g., group of more than 5 people loitering), or movements (e.g., cars moving faster than the speed limit).
In addition, AI can accurately pinpoint the location of event occurrence by displaying the camera that records the event in near real time, i.e., with few-second delays.

Lower security operation cost: By minimizing the number of false positives and human involvement while maximizing ease of use and stability, AI intrusion detection systems significantly decrease the total cost of ownership. Companies can reduce the large security personnel overhead and cost of complicated and expensive legacy PIDS systems. McKinsey Global report in June 2017 shows that proactive AI adopters can realize up to 15% increase in profit margin across various industries.

Artificial Intelligence is undeniably reshaping every business and weaving into every aspect of daily life. In security, legacy systems are giving way to AI-based systems, and the first enterprises to adopt AI-based systems will soon, if not immediately, benefit from such investment.


By Paul Sun, CEO of IronYun and Mai Truong, Marketing Manager of IronYun

Download PDF version

In case you missed it

Preventing Workplace Violence: Considering The Instigator's Perspective
Preventing Workplace Violence: Considering The Instigator's Perspective

A complex set of biological, psychological, sociological, contextual and environmental factors are involved when a perpetrator decides to commit an act of workplace violence. In many cases, the perpetrator doesn’t really want to become violent; rather, they are seeking to achieve an outcome and mistakenly believe violence is their only option. An underused approach to preventing workplace violence is to consider the issue from the perspective of the instigator, to seek to understand their grievances, and to suggest alternative solutions, says James Cawood, President of Factor One Inc. “It’s helpful to consider their perspective at a point of time, and how do I use that information in a way that explores the issues and influences them to seek other means of achieving their goals without violence?” suggests Cawood. Preventing Workplace Violence An underused approach to preventing workplace violence is to consider the issue from the perspective of the instigator Factor One specialises in violence risk management, threat assessment, behavioural analysis, security consulting and investigations. Cawood will present his insights into preventing workplace violence in a session titled “Workplace Violence Interventions: The Instigator’s Perception Matters” during GSX 2018 in Las Vegas, September 23-27. Intervening and seeking to understand the instigator’s viewpoint can direct them away from violence. Often, diffusing a situation can prevent tragedy. Delaying a violent act is a means of prevention, given that the instigator might not reach the same level of stress again. Cawood says several recent examples of workplace violence illustrate the importance of identifying behavioural precursors and intervening. It is difficult to quantify the benefits of such an approach, since no one is keeping statistics on incident that were successfully diverted, he says. Reaching A Mutually Agreeable Solution “Accommodation and appeasement often won’t serve the problem,” says Cawood. “Instead of projecting our needs on what would be effective for us, we must really understand what matters to them and what we are able to do to solve the problem. “It’s about listening and reflecting back to reach a mutual agreement of their perspective of what matters,” he says. “Now we can talk about what’s possible or not. Is there something concrete I can do that is within the rules? Just being heard in depth is a de-escalator of violence.” It’s the same methodology used by hostage negotiators: Listen, reflect back, and come to a mutually agreeable solution. Giving a troubled employee a severance package – money – might not address their underlying complaints For example, giving a troubled employee a severance package – money – might not address their underlying complaints. “We may not have solved the underlying problem as they perceive it,” says Cawood. “They may feel disrespected or picked on. There may be an underlying mental condition, such as paranoia, or a grandiose sense of self-worth, underlying filters that have nothing to do with money.” GSX Networking And Education GSX is the new branding for ASIS International’s trade show, attended by more than 22,000 worldwide security professionals  Global Security Exchange (GSX) is the new branding for ASIS International’s annual conference and trade show, attended by more than 22,000 security professionals from 100-plus countries. Cawood’s session will be September 24 from 2:15 to 3:30 p.m. “My purpose is to hone in on an area of workplace violence that is often ignored,” says Cawood. Cawood started out in law enforcement in the 1970s and transitioned to security in the 1980s. His credentials are typical of the high level of speakers presenting at GSX 2018: He holds a Master’s Degree in Forensic Psychology, and a Doctorate in Psychology, is a Certified Threat Manager (CTM), and has successfully assessed and managed more than 5,000 violence-related cases. He is the former Association President of the Association of Threat Assessment Professionals (ATAP) and currently the Vice-Chair of the Certified Threat Manager program for ATAP. Cawood has written extensively on the topic of violence risk assessment, and co-authored a book, Violence Assessment and Intervention: The Practitioner's Handbook. Cawood has been active in ASIS International since the 1980s and sees value in attending GSX 2018. “People from all over the world are coming and being exposed to a common set of topics to use as jump-off points for additional conversations. People from all types of experiences and exposures will be providing information through those lenses.” Knowledge gained from GSX provides a “real chance to drink from a fire hose” and get a deeper understanding of a range of topics. The relationships and networking are another benefit: “Nothing is more powerful than knowing someone face-to-face,” he adds.

What’s New “on The Edge” Of Security And Video Surveillance Systems?
What’s New “on The Edge” Of Security And Video Surveillance Systems?

By definition, an edge device is an entry point to a network. In the physical security industry, edge devices are the cameras, sensors, access controllers, readers and other equipment that provide information to the IP networks that drive today’s systems. In the Internet of Things (IoT), edge computing refers to an increasing role of edge devices to process data where it is created instead of sending it across a network to a data center or the cloud. In our market, edge computing takes the form of smarter video cameras and other devices that store and/or process data locally. We asked this week’s Expert Panel Roundtable: What’s new “on the edge” of security and video surveillance systems? 

How To Choose The Right Storage Card For Video Surveillance Systems
How To Choose The Right Storage Card For Video Surveillance Systems

With increased demands being placed on safety and security globally, and supported by advancements in IP cameras and 360-degree camera technology, the video surveillance industry is growing steadily. Market research indicates that this worldwide industry is expected to reach an estimated $39.3 billion in revenue by 2023, driven by a CAGR of 9.3 percent from 2018 to 2023. Video surveillance is not just about capturing footage (to review an event or incident when it occurs), but also about data analysis delivering actionable insights that can improve operational efficiencies, better understand customer buying behaviors, or simply just provide added value and intelligence. Growth of Ultra-HD Surveillance To ensure that the quality of the data is good enough to extract the details required to drive these insights, surveillance cameras are technologically evolving as well, not only with expanded capabilities surrounding optical zoom and motion range,4K Ultra HD-compliant networked cameras are expected to grow from 0.4 percent shipped in 2017, to 28 percent in 2021 but also relating to improvements in signal-to-noise (S2N) ratios, light sensitivities (and the minimum illumination needed to produce usable images), wide dynamic ranges (WDR) for varying foreground and background illumination requirements, and of course, higher quality resolutions. As such, 4K Ultra HD-compliant networked cameras are expected to grow from 0.4 percent shipped in 2017, to 28 percent in 2021, representing an astonishing 170 percent growth per year, and will require three to six times the storage space of 1080p video dependent on the compression technology used. Surveillance cameras are typically connected to a networked video recorder (NVR) that acts as a gateway or local server, collecting data from the cameras and running video management software (VMS), as well as analytics. Capturing this data is dependent on the communications path between individual cameras and the NVR. If this connection is lost, whether intentional, unintentional, or a simple malfunction, surveillance video will no longer be captured and the system will cease operations. Therefore, it has become common to use microSD cards in surveillance cameras as a failsafe mechanism. Despite lost connectivity to the NVR, the camera can still record and capture raw footage locally until the network is restored, which in itself, could take a long time depending on maintenance staff or equipment availability, weather conditions, or other unplanned issues. Since microSD cards play a critical role as a failsafe mechanism to ensure service availability, it is important to choose the right card for capturing video footage.  It has become common to use microSD cards in surveillance cameras as a failsafe mechanism if an NVR breaks Key Characteristics Of microSDs There are many different microSD cards to choose from for video capture at the network’s edge, and they range from industrial grade capabilities to commercial or retail grade, and everything in-between. To help make some of these uncertainties a little more certain, here are the key microSD card characteristics for video camera capture. Designed For Surveillance As the market enjoys steady growth, storage vendors want to participate and have done so with a number of repurposed, repackaged, remarketed microSD cards targeted for video surveillance but with not much robustness, performance or capabilities specific to the application. Adding the absence of mean-time between failure (MTBF) specifications to the equation, microSD card reliability is typically a perceived measurement -- measured in hours of operation and relatively vague and hidden under metrics associated with the camera’s resolution and compression ratio. Therefore, when selecting a microSD card for surveillance cams at the edge, the choice should include a vendor that is trusted, has experience and a proven storage portfolio in video surveillance, and in microSD card technologies. Endurance, as it relates to microSD cards, represents the number of rewrites possible before the card can no longer store data correctly  High Endurance Endurance, as it relates to microSD cards, represents the number of rewrites (program/erase cycles) that are possible before the card can no longer store data correctly. The rewrite operation is cyclical whereby a new stream of footage replaces older content by writing over it until the card is full, and the cycle repeats. The higher the endurance, the longer the card will perform before it needs to be replaced. Endurance is also referred to in terabytes written (TBW) or by the number of hours that the card can record continuously (while overwriting data) before a failure will occur. Health Monitoring Health monitoring is a desired capability that not many microSD cards currently support and enables the host system to check when the endurance levels of a card are low and needs to be replaced. Having a card that supports this capability enables system integrators and operators with the ability to perform preemptive maintenance that will help to reduce system failures, as well as associated maintenance costs. Performance To capture continuous streams of raw footage, microSD cards within surveillance cams perform write operations about seventy to ninety percent of the time, whereas reading captured footage is performed about ten to thirty percent. The difference in read/write performance is dependent on whether the card is used in an artificial intelligent (AI) capable camera, or a standard one.   microSD cards deployed within surveillance cameras should support temperature ranges from -25 degrees Celsius to 85 degrees Celsius Finding a card that is write-friendly, and can provide enough bandwidth to properly capture streamed data, and is cost-effective, requires one that falls between fast industrial card capabilities and slower commercial ones. Bandwidth in the range of 50 MB/sec for writes and 80 MB/sec for reads are typical and sufficient for microSD cards deployed within surveillance cameras. Temperature Ranges Lower capacity support of 32GB can provide room to attract the smaller or entry-level video surveillance deployments As microSD cards must be designed for continuous operation in extreme weather conditions and a variety of climates, whether located indoors or out, support for various temperature ranges are another consideration. Given the wide spectrum of temperatures required by the camera makers, microSD cards deployed within surveillance cameras should support temperature ranges from -25 degrees Celsius to 85 degrees Celsius, or in extreme cases, as low as -40 degrees Celsius. Capacity Selecting the right-sized capacity is also very important as there needs to be a minimum level to ensure that there is enough room to hold footage for a number of days or weeks before it is overwritten or the connectivity to the NVR is restored. Though 64GB is considered the capacity sweet spot for microSD cards deployed within surveillance cameras today, lower capacity support of 32GB can provide room to attract the smaller or entry-level video surveillance deployments. In the future, even higher capacities will be important for specific use cases and will potentially become standard capacities as the market evolves. When choosing the right storage microSD card to implement into your video surveillance system, make sure the card is designed specifically for the application – does it include the right levels of endurance and performance to capture continuous streams – can it withstand environmental challenges and wide temperature extremes – will it enable preventative and preemptive maintenance to provide years of service? It is critical for the surveillance system to be able to collect video footage whether the camera is connected to an NVR or is a standalone camera as collecting footage at the base of the surveillance system is the most crucial point of failure. As such, failsafe mechanisms are required to keep the camera recording until the network is restored.