What Keeps Me Up at Night: How Intelligent is My Video Surveillance System?
By Greg Offner
Recently, I attended a private function at Harrah’s Casino in Atlantic City, NJ. Other than the wonderful architecture and engineering features, the lights, sights, sounds and glitter of their Casino, I was genuinely fascinated with the amount of security, both visible and discreet. All I needed to do was look around; at the casino ceilings, at the entries and public movement areas of the resort property to recognize there is a widespread use of video surveillance. I akin one seeing Casino video surveillance cameras to that of seeing a Deer near the highway while driving at night; for every one (camera or deer) you see there are probably 10 you don’t!
I am fortunate to be able to interact with a group of industry experts at PSE. They make me smarter about the SEC universe! These engineering professionals are constantly looking at the trends in security systems and validating newer products and system performance before considering, or even suggesting a product or system for their clients. I value their council and trust their judgement.
Prior to sitting down to pen this article I did some research on Artificial Intelligence (AI), Intelligent Video (IV) surveillance and the networked video surveillance market. What I discovered was astounding! The AI IV market is approaching a $25B market worldwide! WOW. No doubt about it, it has become more obvious that the top video surveillance companies have grown far beyond Casinos and Prisons. The video surveillance market has expanded to provide newer, better solutions in markets like colleges and universities, “smart city” programs and public transportation sectors, to name a few. This growth has allowed AI IV solutions providers to practically quadruple their once robust revenue streams in the last few years.
Hardware Plays Critical Role
A crucial component of any AI-IV system is the use of cameras connected to software analytics, whose DNA is such they can act, or react if you will, if their detectors sense movement, temperature change, even sound that is beyond the ordinary! For example, most people who ride a subway car are static while the train is in motion. They mostly sit or stand still. These latest IV cameras can sense unordinary motion and go live if there is any unusual activity while the train is in motion. They will also go live at station stops to record the safe entry and exit of passengers on the platform or train. Around town, the AI-IV systems have a percentage of the overall surveillance market right now, and their use is growing every day. I will be able to sleep better at night knowing there is a security system is on point.
As I did further research, all the indicators, both market and product performance metrics pointed toward new and better logic enhanced optics, mostly brought on by Drone Technology developed by cutting edge firms like Skydio, Parrot and Yuneec. Indicators are the surveillance systems are gaining traction and working beyond original expectations. IT-oriented companies such as Hikvision, Dahua, Bosch and Axis are accelerating product innovations and enhancements. Their video surveillance system sales are benefiting from the growth of the expanding IP/networked surveillance market.
Artificial Intelligence and Analytics
Already, AI and analytics play a major role in video surveillance, helping end users with security and operational objectives. For example, detecting abnormal behavior, recognizing VIPs like me when I enter a Casino (LOL) or dispatching additional staff when the line is too long at Airport security. Currently, AI is helping with disease prevention and control as well, and this is expected to continue as a trend in our post-pandemic world. Many jurisdictions used CCTV integrated with AI during the pandemic to observe occupancy in public places, and analytics and CCTV integrated with AI is being used to track people movements and behavior and provide immediate data on potential risks, or threats to public and private safety and security. No more tossing and turning? Perhaps.
I share a personal experience with AI that I had during my cancer treatment protocols a couple of years ago. I had a PET scan (Positron Emission Tomography), a medical imaging technique that utilizes radioactive substances to enhance images of tissues and organs to determine if they are normal, or abnormal. Later, the same day I had the scan I left for the airport for a flight to Florida. Once I entered the Terminal, I noticed a uniformed Office grab her collar mounted radio to listen to a call. I knew from experience I had set off the vestibule radioactivity detector. I immediately approached the officer and told her I was aware I had set off the detector and that I had a PET scan that morning. I was in their trusted traveler database and allowed me to use the special screening area and was quickly sent along my way.
The New Normal
With the almost weekly eruption of suspected COVID variants, and the of the potential spread of contamination, using AI to detect virus symptoms using technology like body temperature scans are becoming more common. The fear of contracting the Corona virus contributed to economic and social trepidations, forcing governments to spend billions of dollars attempting to reduce the possibility of spread. This fear factor continues to drive the demand of COVID detection and prevention products at public facilities and especially at our schools and our borders.
Regarding technology segments, the pandemic is causing renewed interest in multiple technology capabilities, ranging from video surveillance and analytics to newer and better identity and access management solutions. These surveillance analytics can include both digital and physical access analysis. These analytics are particularly critical in event management and security operations center (SOC) solutions for them to operating at top effectiveness. Video surveillance, particularly those systems that include integrated analytics solutions, will be critical to enabling customers to return to physical locations of employment or customers, while ensuring that public health protocols can be actively monitored.
The new normal will have analytics solutions that can serve multiple use cases, simply put, counting analytics that can measure distances and number of people within a specified area. These will be some of the top solutions in consideration, as many Owners security teams will be much more particular about the security technology purchases so they can sleep well at night.
Video Technology Changing for Better
It is all about the hardware. What is the value of being a video surveillance company in the AI-related applications and markets? In the AI era, algorithms and computing power from different companies tend to be quite similar. However, when it comes to the reality of usability, the hardware and domain know-how for the installation environment is still very important to provide a better, more social-minded solution.
The hardware manufacturing capability is still developing. Many companies can still offer relatively comprehensive solutions along with leading edge technology innovation. AI companies also need to rely on the companies whose hardware can keep pace with analytics. I look for companies who can provide hardware bundling with their AI algorithm. So more often than not, most sides can benefit from partnerships between each other.
In view of the previous paragraph, let’s say you have camera with artificial intelligence, and it is placed on a light pole or mast in the outdoor rec area of your facility. If you place it at a height that is say 25 feet above the inmates, then it is possible to see the overall population in the yard for people counting, and therefore normal behavior like, loitering, gathering, or participating in sports is easily spotted. However, if you want to see and eavesdrop on potential skullduggeries, you need to place the camera lower on the pole, and perhaps add additional cameras. For any specific purpose in the facility, you have to understand what kinds of cameras you would like to choose and its installation location too. That is why using AI in a specific vertical also contains many different competencies, from hardware, software, installation and service providers.
Security Innovations Every Facility Needs
If you’re a fan of the television programs like CSI you are probably familiar with facial recognition. It is one of the more high-profile applications of intelligent video. The first benefit I can think of is the ability to identify and track in-custody individuals. The process for facial recognition is similar to automatic license number recognition. Facial recognition system draws on a database of faces, such as police mug shots or a drivers license photo database. Getting that data is often the biggest challenge in deploying the system. Once that database is accessible, facial recognition undergoes a multi-step process to match the video image to a stored picture. Corrections officers and Courts Security staff can use the technology to alert them when certain people of interest are entering or leaving their facility. Courts and Correctional Facilities can use the technology to enhance access control, allowing only certain individuals to enter specific areas. Border agents improve checkpoint control by augmenting manual passport checking with automatic searches for individuals of interest.
I want my AI to have the ability to be used for Life Safety functions. Fire and smoke detection systems search for visual cues of fire and/or smoke in the video stream. They can be set to alarm as soon as a flame is visible in a room-or when reflected firelight is detected from an obstructed view-without having to wait for ambient smoke to appear or a puff of smoke reaches the ceiling or in-line duct detector. This gives facilities managers a much earlier warning of problems than a simple smoke detector. However, there may be fire codes which prevent video identification as the sole system for fire detection which has resulted in limited widespread adoption of the technology.
Beyond that decision, there are some technical factors to consider that will help you improve the accuracy of your facial recognition application, such as where you place cameras, and what analytical functions you want them to perform. To my knowledge, no IV application is infallible. Should you decide to deploy AI technology, you need to have realistic expectations. Reaching a 99 percent accuracy rate can be expensive, far beyond that of a traditional IP CCTV systems but sleep tight, today it is not impossible.
The Big Bang Theory
The Intelligent Video surveillance industry is growing, and the use of IP Cameras and IV analytics is changing the way we protect our environment. The market for IP/networked video surveillance cameras and object-based video surveillance systems had dramatic growth early in the 21st Century. In retrospect, the growth of the CCTV based video surveillance equipment market, including cameras/DVR and other video surveillance systems which have been in use for a few years has diminished in favor of smarter, better technologies. However, according to a few large manufacturers analog systems are still very much in demand.
Frankly, I believe AI IV video analytics is in its teenager stage of development and for now, both analog and IP cameras are being used with the IV technology. One thing is certain, object based Intelligent Video is the next big bang in the prison, jail and courts security industry. I for one hope the big bang doesn’t wake me up!
Gregory J. Offner, CCM, a Correctional News Editorial Advisory Board member and regular contributor to the publication, serves Professional Systems Engineering, LLC (PSE) as a senior associate.
Editor’s Note: This expert column originally appeared in the May/June 2022 issue of Correctional News.