Learn to Walk Before Running

AI technology has been a pop culture fixture for generations

• By Alex Walthers
• Mar 02, 2021

It doesn’t feel right to call artificial intelligence (AI) a “new technology.” After all, intelligent machines have been a pop culture fixture for generations, and although Hollywood’s depiction of AI isn’t always the most realistic, it has — for better or worse — contributed to the technology’s considerable hype. Today, manufacturers and developers have been working diligently behind the scenes to bring practical AI closer to reality. Through years of baby steps and continuous improvements, they have established a solid foundation for the future of AI—and across countless industries, businesses are exploring the advantages the technology can provide.

This race to embrace AI has, unfortunately, led to some confusion in the market. After all, what is “AI,” really? What can it actually do, and how is it most effectively used? In the security industry, it is particularly important to beware false claims. AI has greatly enhanced the capabilities of many cameras and other sensors, but it is critical to understand that implementation of the technology is still in its relatively early stages—and much of the AI hype remains speculation. Rather than looking to the future, today’s businesses should focus on what today’s AI is actually capable of.

UNDERSTANDING AI AND ITS EVOLUTION

The analytics enabled by AI and its various subsets are significant, but misconceptions abound. Hollywood, once again, is not innocent here: how many people watched Tom Cruise investigate “pre-crime” in Minority Report or saw Jack Bauer “enhance” blurry security footage to reveal a crystal-clear license plate on 24? On CSI, you can expect to see someone identified by their reflection in a hubcap roughly once an episode. This level of hype for the future of analytics persisted for years, but the fact that reality never quite seemed to measure up led many to view the technology as a disappointment. Today, we have in many ways caught up to those prior expectations — Jack Bauer would be thrilled with today’s advancements in video quality — but new misconceptions persist.

As with most things, change in the artificial intelligence world has been incremental. Video analytics is a great example of this. In the past, traditional video analytics looked primarily at the change in pixels within an image. Want to know whether someone crosses a designated line? No problem: draw a line and measure the pixels across that line. Unfortunately, that type of analysis can’t take into account whether what is crossing the line is a person, a dog, a cat, or even just a spider on the camera lens. All it knows is that pixels are changing.

Advancements in analytics have allowed us to take into account much more than the movement of individual pixels. Much of today’s AI is based on implementing and comparing different models, such as a database of images. The better the data set, the more effective the AI. This has enabled a wide range of impressive improvements, such as the ability to identify a shadow cast at a certain time of day, rather than mistaking it for a new object in the area. This has helped dramatically reduce alarm fatigue, which remains a serious issue in the security space. Over time, these incremental improvements result in significant achievements that provide practical solutions with great accuracy.

TRAINING AI TECHNOLOGY IS KEY

If an AI is only as good as its data set, then improving those data sets is of paramount importance. While you might expect critical industries like healthcare or security to have the most cutting-edge technology, the truth is that most analytics programs are used in non-critical situations first in order to refine them for more mission-critical uses.

Retail. If a shelf is stocked with the same products all the time, the AI can be trained to recognize those products by comparing against other images. When that product is depleted, it can alert the store that it needs to be restocked or reordered.

Manufacturing. Routinely deployed, a camera, a camera watching the assembly line at a cookie factory, for instance, might be trained to identify when a cookie failed to receive its cream filling. It is something important to that particular company, but it isn’t safety critical. This provides a clearly defined benefit to the user, but also provides long-term benefits by fine-tuning the technology. Those same technologies, now more refined, can then be applied to security and other critical applications.

Using AI in these environments helps build the necessary models, adjust to different situations, and identify what is important and what is not. This is also why many new AI solutions are essentially “next step” improvements on what already exists. It is easier for a company to improve upon an existing solution if they have thousands of images (and years of experience) working on that same technology. Access to this type of training data can help power incremental improvements to the technology, underscoring why AI adoption requires a more thoughtful and measured approach.

To put it another way, think of how long it takes a doctor to learn to identify signs of one specific disease. An ophthalmologist might look at tens of thousands of images of eyes showcasing the exact indicators to look for. AI, on the other hand, is trained on data sets that aren’t nearly as specialized. After all, developers can’t just go to a local retailer and collect random video that fits their needs. They have to rely on more abstract videos of cars, shadows and things crossing a line in space. While some have experimented with creating synthetic training data, there is significant risk involved: what if one element is wrong? One small mistake can ruin an entire training program. As a result, training AI in noncritical applications remains the most effective way to refine the technology.

AI’S IMPACT ON SECURITY AND OPERATIONAL EFFICIENCY

Once technology has been proven in noncritical areas, it can be deployed in more critical applications. As previously mentioned, one of the biggest challenges in the security space is alarm fatigue, or the issue of too many false alarms. If sensors cannot tell the difference between a trespasser entering a building and a shadow moving with the sun, security teams will be deluged with false alarms, potentially drowning out actual security incidents amid all the noise. Forcing security staff to sift through dozens of alerts is time consuming and inefficient, and increases the likelihood that a genuine threat will slip through the cracks.

Today’s analytics are not only better able to differentiate between true security incidents and false alarms, but can be programmed to trigger when a given event occurs. That event might be a person in an area they aren’t supposed to be, too many people standing in line at a register or even someone not wearing a mask.

If detected, an alert can be issued to the appropriate staff and a designated response can be triggered, such as a light turning on, an audio message playing, or even a human interaction. From a security standpoint, this works as a highly effective deterrent: shine a light on a bad guy in an area they’re not supposed to be, and they feel busted. An audio alert that lets them know they’ve been seen will, more often than not, lead them to beat a hasty retreat, foiling whatever plans they may have had.

There is also a business intelligence aspect to artificial intelligence. Today’s advanced analytics can combine and analyze data in new and innovative ways, leading to insights that can greatly increase operational efficiency.

A weekly data report for a retail store might include data correlating sales numbers to a recent marketing promotion, or an analysis on which direction customers are entering the store from, or which direction they are heading once they enter. It can even focus on how customers are dressed, identify where choke points are within the store, or track how many customers enter and exit the store without making a purchase. The ability to track data the ranges from what color customers are wearing to whether they arrive via car, SUV, or public transportation can help stores make more intelligent decisions on how to present themselves in the future.

There are also areas where security and business intelligence combine. To return to the assembly line example, cameras can be trained to observe the various steps in the manufacturing process and understand the actions involved, enabling them to identify mistakes or inefficiencies. This can not only save the business money down the line, but potentially identify manufacturing errors before they can become a liability to the company.

It would be a mistake to overlook the value of audio analytics in addition to video. Today’s audio solutions can be programmed to detect certain sounds, such as raised voice, breaking glass or gunshots, that might indicate a security incident, even in total darkness. Audio can also add context: it can sometimes be hard to tell from video alone whether two friends are joking around or are actually about to fight. It isn’t always easy for a machine to make the right call based on audio or video alone, but both together can provide a human with the information they need to decide what is actually happening. Audio solutions are also an important option to have in settings where a camera would be inappropriate, such as a school bathroom or other private area.

THE FUTURE OF AI IS BRIGHT— BUT IT’S IMPORTANT TO BE REALISTIC

Artificial intelligence and its subsets, machine learning and deep learning, have enabled the creation of more powerful analytics than ever, and the ability to train this technology in non-critical environments has been essential for the technology’s continued development. What’s more, the development of advanced chipsets and more powerful processing units has made processing and analyzing data at the network edge both actionable and efficient. These deep learning processing units will also help to propel the development of more sophisticated and accurate algorithms.

But it’s important to remember that while AI is excellent at simple, repeatable functions, humans are significantly better at interpreting that data and deciding how to respond. For this reason, it can be best to think of AI not as artificial intelligence, but as “augmented” intelligence, providing valuable new information to help human beings make better, more informed decisions. As AI continues to evolve, so will its uses—but understanding the technology’s current uses and limitations will ultimately help businesses get the most out of today’s artificial intelligence solutions.

This article originally appeared in the March 2021 issue of Security Today.