In this Knowledge Partner article, Robert Prady CPP, PSP, CSP, Area Technical Manager, West Business Area, Axis Communications explains how network video and AI-based analytics are improving security, adding operational efficiencies and enhancing customer experiences.

Casinos have evolved from humble gambling houses and riverboats to today’s mega-resorts. But, their security risks have grown with the; luckily, casino security practices have kept pace.

Today, it’s standard for casinos to conduct formal risk assessments, develop mitigation plans, implement management protocols and utilise modern technologies. Chief among those technologies is network video surveillance – arguably the most important tool in the toolbelt. So, where does casino security go from here? What do the next ten years look like? Modern network video cameras armed with powerful processors and AI-based analytics hold the key.

Shifting from monitoring to investigation with analytics

Video analytics are nothing new in the world of surveillance. In fact, they’ve been in existence for the last 20 years. However, it took today’s powerful processing units and advanced software – both residing directly on the camera – to reach a new level where technology can automatically perform a large part of an investigator’s job. As a result, more and more casinos are shifting from a live monitoring posture to a forensic investigation stance.

In the past, it took the majority of a casino’s investigators to “live monitor” operations via a wall of video screens while only one or two investigators were available to conduct reviews and create reports after the fact. Today, the tables have turned for the better, live monitoring only requires one or two operators and the rest of the staff are able to conduct post-incident investigations. This is a side effect of improved data collection abilities. With more devices in the field collecting data, this trend will continue, and at some point, live monitoring may cease to exist as a standard operating procedure and may only be reserved for active incidents.

Reducing workload through object recognition capabilities

Currently, object recognition video analytics are being used to tag and classify objects within the camera view, such as a car, a person or a bus as well as the color of those objects. That classification data is stored as metadata within the video stream and that metadata can be searched as stored video. For example, we no longer need to search only on camera time and date – we are now able to search for a person in the red shirt on all cameras. 

This capability dramatically reduces the workload for the investigator and can provide additional information such as an individual’s known associates quickly. Many times, this information can be accessed proactively as an incident is still unfolding.

The value of object recognition and classification is just being realized, and adopted, by the casino industry. In the next ten years, this technology will be commonplace, especially as the cost of the technology comes down. Object recognition and classification is achieved by what is known as “machine vision.” In short, machine vision is the use of camera hardware and software to help cameras see, analyze and act. This is accomplished by scanning individual snapshots of video as they are being captured by the camera. As this technology matures, we will be able to identify and classify more objects beyond people and vehicles including things like weapons, smoke and other items that we would want to bring to a human’s attention. 

Gaining greater insight as technology advances

The most significant improvement expected with this technology over the next decade will be the ability to identify objects in relation to other objects and compare to neighboring frames. So, for example, in addition to identifying a handgun, we’ll be able to tell if the handgun is in a holster or in someone’s hand. In other instances, we’ll be able to tell if a patron slipped and fell or just crouched to pick up something that they dropped on the ground.

If a person is captured running inside the casino, we can quickly see a ten second clip just before the person started running in order to determine if an incident took place beforehand. These are only a few examples of common capabilities and uses that will come to fruition in the next ten years.

Game protection – which is the loss prevention and detection of scams and advantaged players in the casino – will also benefit from the advancement of AI and video analytics. Currently, there are analytics that can recognize the pips, or symbols, on cards in order to identify win/loss per hand and even record the pace of the game. In the future these analytics will be able to determine player betting patterns, identify advantaged players and possibly even automatically rate the individual’s play and link that to a specific player via facial recognition.        

Stewarding the increase of IoT devices and resultant data

As IoT devices become more intelligent, there’s opportunity and valid rationale for surveillance departments to be the custodian of all this data. Not just video data but license plate data, access control data, facial recognition data, along with real time information sharing with local law enforcement and emergency services and, as a result, we’ll see improved outcomes. All of this data will help surveillance departments become more efficient – better protecting guests, employees and assets with greater expediency and accuracy. 

Stewarding all of this data may sound like additional work, but it truly makes sense to centralize this vital data and entrust it to a qualified security team. That said, as technology advances and data becomes more abundant, it’s important to involve, and train, technical staff as well. Essentially, casinos must transform traditional surveillance technicians into surveillance systems engineers.

Casinos must focus on evolving their responsibilities beyond just installing cameras and maintaining recorders to include managing and administrating all of the data being received by intelligent IoT devices. No longer is IoT the sole domain of the IT department—expertise and responsibility must be shared with the surveillance department. This transition will gradually occur over the next decade, which means planning and development must occur now.

Limiting losses, increasing efficiency, enhancing experiences

In the next ten years the surveillance landscape will change significantly. Casinos will have modular cameras inserted into locations that were previously never considered including gaming tables, ticket redemption machines, slot machines and any other place our imaginations can conceive.  Analytics like license plate recognition (LPR) and facial recognition systems will be commonplace as their accuracy accelerates. What’s more, these systems will not only link to repeat offender information for security purposes, but they will also link to valued player information to provide better service and enhanced customer experience – identifying VIP patrons and ensuring that staff are aware of their presence in order to render a higher level of guest service. 

Systems will feed a database where desired and undesired patrons activity can be reconciled against players club accounts thereby reducing fraud and increasing the guest experience, ultimately increasing casino efficiency and profitability. Modern IP cameras and other devices will now capture video for recording as well as multiple data points for storage.

With increasing amounts of data being generated, file compression technology and storage capabilities will become even more important considerations. Regardless, the data will be worth its weight given its value in incident investigations. Despite storage needs, the connectivity of the IoT environment will allow surveillance rooms to be consolidated. Increasingly, multiple casinos will be monitored from one location and even allow surveillance investigators to work at home, if allowed by regulatory agencies. 

Capitalizing on future opportunities and navigating challenges

Since the inception of network cameras and video analytics, many prior predictions are being realized and some have come to pass – even remote casino monitoring has been allowed by some select jurisdictions during the pandemic. If the pandemic has taught us nothing else, it is that we must be agile and have the ability to adapt to the ever changing business landscape quickly. 

There’s no doubt that network surveillance solutions and AI-based video analytics offer great potential to improve security and enhance operations. Their benefits will be more fully realized in the next decade. However, as with all new technology, costs and regulation can be obstacles so we must learn to navigate them. That said, it’s a universal truth that the cost of technology will go down as it matures. More importantly, casinos will be able to quantify a ROI as these technologies reduce losses and increase efficiencies.  

As far as regulations, many rules on the books across the country are still based in analog technology. As an industry, it’s important for casinos to work with regulators to update these antiquated regulations. Ideally casino operators and industry associations, not vendors, should advocate for their needs thereby driving the future of casino security and operations and the advancement of technology that helps to ensure them.  

This article was originally published in the December edition of Security Journal Americas. To read your FREE digital edition, click here.