While Intelligent Video Analytics advances have expanded beyond basic security applications to wide variety of use cases across many different industries, supporting new business opportunities, customer experiences and service objectives. The casino gaming industry in considered leaps and bounds ahead of most industries. My recent research in which I spent eight months actively playing and discovering IoT and intelligent video analytic applications.
With this hand on approach in Las Vegas, I have gathered a unique perspective of how video analytics is used in a few of the top properties within the convention, hotel, transportation and gaming industry and in Las Vegas.
The primary reasons for video cameras in the casinos are for security and traffic data. However, casinos believe that more value can be obtained by transforming the unstructured data (such as footage that a guest has stopped playing a gaming machine) into structured data events that can then be correlated with data coming from existing business systems.
For example, using video analytics the footage of a guest leaving a slot machine can provide insight in the form of metadata — such as the machine number used by the guest, the time stamp when the guest left the machine, and possibly even a determination of whether the guest came back to the same machine with a drink in hand or how many drinks a guest has consumed.
This metadata can now be compared against data in the existing enterprise data warehouse (EDW) that came from the gaming machines and guest tracking systems. For example, the gaming systems keep track of how much time and money the guest used at that specific machine and at that specific time (because most players “swipe in” with their loyalty card). These tracking systems has records on what the guest spends his money on at the entire casino.
The combination of these trackin sources of information can provide business insight on why players leave machines, which guest segments leave and for what reasons — and do they come back. Personalized offers could eventually be presented to guests as they play.
- The same camera infrastructure can be used for many functions: Previously, casinos used its analog cameras for security monitoring purposes, but it was unable to provide a more holistic view of all inter-related casino/floor operations and customer activities. When cameras become IP-enabled, video analytics can be configured for many uses, such as customer counting, security incident detection, traffic flow and customer action detection. This multifunction capability is a key benefit in delivering a favorable ROI, because the capital investment in cameras, storage and infrastructure can be offset by business benefits in many areas.
- Increased situational awareness for security and customer experience issues: Real-time video event detection provides capabilities and benefits that are rarely possible in analog video systems. Current video analytics software can detect time, event, forensic analysis, identification, behavior analysis and situation awareness. Organizations looking to automate video surveillance systems should IP-enable their systems to allow for more real-time analysis and remote monitoring, and to apply additional security applications.
To gain these benefits, casinos have to overcome some of these challenges:
- As networked IP-based video surveillance systems evolve, they will become more IT-centric, with cameras using IP addresses that can be centrally controlled with a variety of software applications requiring more participation from IT departments.
- Historically, physical security managers opt for more analog cameras, cables and localized digital video recorders (DVRs) to offload storage and, at some point, this redundant process — without IP-based filtering software applications and/or a centralized IT infrastructure. This results in a state of information overload, because proprietary analog systems are not interoperable.
Now CIOs & Security Operations Managers determined that the technology and economics favor centralizing IP-based video surveillance by using an IT infrastructure to filter all content and create a searchable video content database to improve situational responses.
Video requires careful architecture planning, automation and next-generation analytics to minimize storage, bandwidth and computing power.
Migrating from installed analog-based camera systems and recording devices raises certain challenges that must be answered before you begin (for example, supporting multi-vendor devices). The objective is to avoid simultaneously operating analog and IP systems, because this proves to be a cost and management challenge. Casino CIO’s need to develop a multi-phase migration strategy that consisted of the following:
- Perform a needs assessment for IP-based/networked video surveillance systems. Compared to analog cameras, networked IP-based cameras are more capable of zooming, refocusing and adjusting to lighting conditions, and displaying results in a more unified manner.
- Determine a life cycle (end of life) replacement road map to replace analog-based DVRs, network video recorders (NVRs), cameras and switches with an IP-based system. Such an IP-based system should be server-based and connected to a storage area network to remove the potential for installing localized hybrid DVRs/NVRs. This type of architecture supports HD and megapixel bandwidth rates that allow for remote viewing and tracking of suspicious behaviors/movements with analytical-enabled cameras.
Many Casino CIO’s are dealing with different technology systems to evolve toward performing a singular task that now shares IT infrastructure that can inter-operate with multiple IP-systems. CIOs need to have a data strategy for video.
Video is arguably the most resource-heavy data of all the data types an enterprise will encounter. Thus, it is vital that CIOs and their teams have a clear handle on these key questions:
- What resolution/compression levels are adequate? Generally, better quality leads to better insight — but at greater cost. Chickasaw Nation Division of Commerce sought a balance between quality and cost. For example, it made its decision on the video quality based on the function of the camera, because certain areas need higher resolution cameras than others. It is also looking at tiered levels for frame-per-second (fps) quality (for example, one area may capture at 30 fps; whereas, a less important area might only record at 15 fps and be triggered only when there is motion detected).
- What video data should be kept? At this stage, Chickasaw Nation Division of Commerce is still tweaking what makes sense for it. It has figured out some of the easy pieces, such as saving event metadata (for example, time stamp, where the footage was taken, and details about who is in the picture). It only keeps full footage of video when it is required for forensics for security incidents. Chickasaw Nation Division of Commerce is pursuing better video analytics software that may give it access to even more metadata.
Impact: When enterprise assets become part of the Internet, they can become the CIO’s responsibility — creating both opportunity and radical changes to IT resources
As trends in operational technology and the Internet of Things play out, a large and varied number of new Internet-capable devices are being added to enterprise networks and the Internet. When this occurs, there is a good possibility that the new devices and their information streams will become part of IT’s responsibilities — CIOs need to be ready for thisresponsibilities was the video camera infrastructure. The trend in video technology has been to move from an analog closed-circuit technology to an IP addressable technology that makes cameras and their video footage accessible via the Internet.
The result of this seemingly small but important Internet enablement of cameras is that the CIO of Chickasaw Nation Division of Commerce now became responsible for the video footage data, which put a massive challenge in his responsibility: 3PBs moving to 5PBs in the next 12 months (representing a rolling 14 days of footage from 4,000 cameras). The impact was huge. Chickasaw Nation Division of Commerce’s previous data center scale was only 200TBs — mainly representing data from business systems and gaming machines. Because of privacy and security regulations in the U.S. gaming industry, this footage could not be stored or processed in public clouds, so the CIO had to build a data center to deal with the volume. Many tools and techniques to manage this scale of data had to be employed, including server virtualization, video compression algorithms, and a strategy for determining what to keep and what to purge, such as keeping metadata of events instead of the video events where it makes sense.
Even though this research looks at a casino’s use of video, many parallels exist across all industries. For example, CIOs whose enterprises go from ISDN to IP-based video conferencing will have to deal with similar new responsibilities. Fortunately, the Chickasaw Nation Division of Commerce CIO already had experience dealing with other non-PC/mobile devices being part of his network — in this case, casino gaming machines. So, there was no culture shock of having to deal with the realization that the Internet of Things could add all kinds of new devices to his network and data center.
- CIOs need to anticipate which Internet of Things devices and data streams will become part of their responsibility.
The magnitude of impact can be quite substantial in scale, type of device and type of data. CIOs need to have discussions with COOs, facilities leads, chief marketing officers and any other business leads responsible for technology that will or has become Internet-enabled. Issues of governance, funding, ownership and life cycle management must be explicitly determined as enterprise things (such as IP-enabled cameras) and consumer things (such as connected cars or home automation) become part of the CIO’s responsibilities. CIOs need to formulate strategies by asking questions, such as:
- Does the CIO own the device or just the data stream?
- Who is responsible for ensuring that the devices are secure when they are accessed and managed on the corporate networks and, more importantly, via the public Internet?
- How can the CIO help the business discover innovations and business improvements, as enterprise assets and consumer devices become part of the Internet of Things?