My (Computer) Vision for the Twenties
This is the first of a three-part series (a “Willogy”) where I explore the opportunities for casino surveillance operations over the next decade.
Part One: Artificial Intelligence
It’s January 2020, it’s a new year…hell, it’s a new decade! The gyms are full of new members, Amazon trucks are delivering diet books to houses across the nation, people are re-activating their annual resolutions with an optimistic new vigor and businessmen are either plotting world dominance or contemplating ways to stay afloat.
I always feel sorry for industry journalists and bloggers at this time of the year who for some reason feel obligated to give their readers crystal ball predictions for the upcoming year. They must dread it. How are they supposed to know what’s going to happen in the future? Most play it safe and write optimistic predictions based on trends from the previous year, trying to be as nice as possible to their advertisers. For once I would like to see an article predicting that initiatives like self-exclusion programs, skill-based gaming, enticing more millennials to gamble and casino social media directors are bull s**t and will be gone by the end of the year.
I won’t be that guy. I will, however, attempt to interpret what I have learned over the last twenty years in the field of casino surveillance and game protection, peel back the layers of the onion and point out where I feel casinos continue to drag their feet in game protection, security and compliance issues. I will give you my take on problems that exist and give my views on where I think surveillance should head in the future. I may piss some people off. Spoiler alert: I think surveillance should change…not a little bit, a lot.
When I look back over the first score of the 21st century probably the biggest change I’ve seen in casino surveillance is the transition from analog to digital technology. Gone are VHS video tapes, substandard resolution, 12 inch camera domes and handwritten reports. Digital technology has shaped the way we do things. It has allowed us to complete daily tasks and requests quicker, easier and more accurately. But has anything else really changed in surveillance? Is surveillance getting better results now? Are they catching more bad guys? Are casinos better protected from threats than they were 20 years ago? It’s arguable.
There’s no argument that in everyday life technology is progressing at lightning speeds, for those that can afford it. There’s more information, more convenience, more screens, prettier images and our lives have been enriched by the many new digital friends we now have. Our friends like us even though we haven’t actually met them.
But has there been an increased return on new surveillance technology over the last twenty years? Not really. It just looks better and makes our job easier. I think it would be an interesting exercise for surveillance directors to compare how many bad guys they have caught each year over the last twenty years and see if they’re catching more with their upgraded technology. I wonder if the increased amount of investigations and inquiries could have been handled with the same technology they had last century?
Like our cars, casino surveillance technology looks better, it’s more comfortable and has more bells and whistles but at the end of the day (whether old or new) it still achieves the same objective of getting you from A to B. It’s really all about the driver behind the wheel. Can cars drive themselves? Do they know how to get to where you need to go?…Oh yeah, they can. Point taken - I need a new car; so do casino surveillance system designers and directors.
Casino Surveillance Technology: What I Have Learned
The bigger question we need to ask moving forward is how do we get more value from camera technology? How do we design an all-encompassing video system that can be integrated into our entire property so that all managers will be provided with real-time actionable intelligence so that they can perform their functions better and contribute more to the organization’s goals? How do we effectively and efficiently protect our casino from threats while at the same time use video to run a dynamic business and increase profitability? How can video “show me the money”!?
Before I give you my take on the direction I feel we need to go in terms of surveillance technology, here are four hurdles the casino industry needs to get over first.
Casinos are cheap. Sorry vendors, I know the thought of scoring that prized casino trophy gets y’all looking for a bigger house on Zillow but take it from the ones that have been in this business for a while, casinos are tight when it comes to surveillance equipment. Sales cycles are 1-2 years and casinos don’t like paying more than they have to for security. Oceans Eleven casino security - a fantasy. Most executives see surveillance as a necessary evil but a drain on profits. Quite frankly I get it. If surveillance isn’t bringing home a piece of meat every now and then, they are out of sight, out of mind. That is until the accountants read the monthly bill for surveillance. It took a long time for the industry to fork out for new digital video systems. Some even paid sub-contractors to make VHS recorders after major suppliers stopped selling them so they could delay the cost of transitioning to digital recording. Even now a lot of them are using a hybrid analog/digital system.
How to jump the hurdle? Develop a surveillance department that adds value, work for a casino company that is technologically led and/or is making lots of money.
Regulators make changing things difficult. Especially big organizational changes. Casinos operate under a stringent regulatory environment. Trying to change what’s written in the book is like trying to pry a joey from its mama kangaroos pouch. They like to sit back on their tails and kick their big feet at you until you’re laying helpless battered and bruised (but not bleeding). Then they say no. It’s bureaucratic waterboarding.
How to jump the hurdle? You specify the benefits to the state and build a relationship of trust.
Integrating different technologies is challenging. A lot of stuff has to come together to make big game-changing systems work. The major concern and buying considerations I have repeatedly heard from surveillance directors when buying systems over the last 20 years is reliability. Will their integrated system work…and stay working? I would have thought in this day and age a buyers main consideration would be use-ability and features but the main priority is that it doesn’t break down. Actually, it’s a regulated requirement that can shut a casino down. Unfortunately, I continue to hear about installation disasters in casinos all over the country. The blame game is played between the casinos, integrators and manufacturers. For me most disasters are brought on by either a weak link of knowledge somewhere within the three parties, not everyone being on the same page or taking short-cuts to save money or time. I question the expertise of a number of integrators but not as much as I question the digital technical knowledge of a number of casino surveillance directors.
How to jump the hurdle? Casinos pay for independent technology experts to oversee system integration.
Casino managers from the old school need to enroll in the new school. It’s a culture thing. Casinos are notorious for creating department silos. Surveillance directors are probably the biggest culprits (closely followed by casino managers). They’re not comfortable with change. Hell, a lot don’t even question why it is they do what they do. An “us and them” relationship has existed pretty much since casinos first installed surveillance operations in the late 70s. Unfortunately, the concept of employees secretly looking over other employees shoulders sometimes makes for an adversarial relationship. The power of video technology cannot be realized to its fullest potential for a business if the intelligence gathered is only confined for the use of one department and one purpose only.
How to jump the hurdle? The surveillance department stops acting like big brother and starts acting more like a sugar daddy (to a point).
The Future of Surveillance is Computer Vision
The world is on the cusp of an artificial intelligence revolution. As a career casino surveillance guy I’m very excited about the potential of a field of artificial intelligence called “Computer Vision.” Computer vision trains computers by using deep learning models to interpret and understand digital images from cameras and videos so it can accurately identify and classify objects — and then react to what they “see.”
This technology could change the way surveillance is conducted in casinos. For decades surveillance operators have been asked to sit in a chair and watch the casino by clicking through hundreds of cameras in the hope that they will see something untoward. The odds of picking up anything is stacked against surveillance people and success is usually dependent on how smart surveillance operators are.
Imagine if all our cameras were smart, programmed to monitor the games by themselves. There’s an old saying in surveillance that cameras don’t catch people, people catch people. Smart cameras will catch people. I’m talking about total automation of monitoring casino games. No humans required to monitor the game. No floor supervisors. No surveillance. A driverless game protection car.
How would that work? A player cheats - ALERT! A player is counting - ALERT! A game rule or procedure is breached - ALERT! The game is not being dealt efficiently - ALERT! A dispute at the table - ALERT! A suspicious transaction - ALERT. A person of interest - ALERT! A player is winning big- ALERT! A dealer steals - ALERT!
The benefits of computer vision will not only include protection but profitability. Once and for all casinos will be able to accurately determine the worth of every individual player and comp them appropriately. By being able to compute every bet placed from a camera we will be able to get rid of the archaic practice of manually rating (and over rating) players. Game pace and table utilization information of every table will also be able to relayed to management in real-time.
It sounds like pie in the sky but actually companies have been working on computer vision technologies for the sole purpose of monitoring casino games for a while. Challenges still exist, mainly how a camera can identify the amount of a player’s chips. In my opinion the answer is simple, get rid of the chips. However, as I alluded to earlier, a lot of casino managers are more comfortable in their old school and are reluctant to digitize table games. They feel chips are an important part of the player experience. I see digitizing the wagers and interfacing with computer vision technology as a catalyst to taking away the guesswork and making games more efficient, protected and profitable in the future.
Caution: Snake Oil Salesman Approaching
Over the next decade there will be a myriad of start-up companies, fresh out the offices of Silicon Valley’s venture capitalists that will claim to have an innovative state of the art AI product just for us. It’s already begun. They’ll have fancy techie names and claim to be world leaders in AI. They’ll have investors with recognizable names and will find their way into our lives through partnering with reputable long term casino vendors who we have existing relationships with. They’ll have spacey websites and their marketing spin doctors will send out press releases with big hairy audacious claims. Don’t fall for it. Be careful. A year ago they never knew anything about casinos - now they’re experts.
It is important that if we are to effectively incorporate a computer vision system and change the casino world, knowledgeable casino operators and analyzers should be involved in determining what we need to see. It will start with studying the individual intricacies of a game and using past history and analysis to determine what is flat-line normal and what is not. It will take a collaboration of operational and technical experts to work out what action could be seen on a game by a computer that could indicate a likely threat.
The trouble with buying off the shelf software products is that when it comes to casinos, one size doesn’t usually fit all. Casinos work under different regulatory requirements and management team priorities vary from property to property. There is no international or even national standard when it comes to casino surveillance. It will take a partnership between knowledgable casino people and patient programmers. But it will be worth it!
End of the World as We Know It
So what will this mean for surveillance people? As I mentioned earlier I don’t have crystal balls. I just know that ever since I have been a surveillance guy I have always wondered about how much stuff we don’t catch. I don’t mean to sound like a glass half empty guy, I try not to. In fact I think surveillance people do a great job given the task and their limited resources. I’m just saying let’s keep it real, the odds of finding that needle in the haystack are getting worse. We can’t keep asking a few surveillance operators to sit in a room and effectively watch everything that goes on in a casino and respond to the ever-increasing management requests at the same time. We’re kidding ourselves. We either get more people or turn to technology.
In fact, what I’m suggesting is computer vision could play a bigger role in the overall running of every aspect of a casino operation in the future. Assertive surveillance directors potentially could lift their teams out from behind the dark room of secrecy to become almost god-like within their organizations by providing intelligence to keep their casinos not only protected but ahead of their competition in terms of market share and profitability. To that end, I think it’s time to ditch the Surveillance name and change it to Business Intelligence.
What will happen to our surveillance brothers and sisters when computer vision is implemented? Personally I think the core purpose of casino surveillance is to assist in maintaining the integrity of the games but I would like to see surveillance expand their role to do more investigations, audits and business analysis. I see an opportunity to become the casino’s internal “Business Insider” magazine. A central intelligence agency that provides the organization with valuable intel.
Of course, to achieve these goals there would have to be a change in the skill set of staff. This will require people with higher education or more specialized training. Essentially individuals would be required to be experts who will be focused on the business and results. Game protection will be left to computer vision technology and a special response team. I see surveillance (intelligence) teams becoming smaller and more specialized.
I go back to my unanswered career-long question: what don’t we catch? It might just turn out that after computer vision cameras are installed we discover that we need twice as many people in surveillance to deal with all the alerts. We may discover that we can’t handle the truth.