Source – facilityexecutive.com
In the world of physical security, now driven so heavily by cutting-edge technology, there is information coming in from a multitude of data points. From the control room, security personnel and facility professionals are on the receiving end of massive amounts of information from various sensors and systems, from reports from access control systems on current badge swipes, video clips from surveillance cameras, and reports on vehicles — sorted by color, make or model — that entered parking lots in the last 24 hours.
So large and complex a data pool that neither individuals nor everyday computing systems can handle it, big data, as it is often referred to, is an aggregation and analysis of these data streams to find anomalies that people and many machines wouldn’t be able to find on their own. It can even identify problems an enterprise may not even know it had. The key is finding a way to better harness and understand it.
Many companies such as Google, Amazon, and Facebook are already mining big data based on people’s actions online and with social media, using sophisticated algorithms that automate the collection, analysis, and related actions based on the results of those efforts. These companies see the tremendous value in this data differently than most businesses currently do.
With its increasingly sophisticated management platforms that continually gather data, the security industry has the opportunity to leverage this information as well. Being able to harness data from physical security systems and break it down into more usable packets can not only enhance its use for forensic security purposes, but also to offer some predictive tools so users can make proactive decisions.
Every sensor within a system has data to share. Whether it’s fire, intrusion, or access control, the information can be used for its own security-related purposes, but it also can be used to indicate patterns or key learnings. This means going beyond just gathering and recording data from a system to find out, for example, the peak hours of operation for an access control system and receiving instead key learning that can give predictions about future usage and needs.
Or consider this scenario: a customer in a metro/downtown area in a large commercial high-rise has an access control system that integrates with the turnstiles in the ground floor lobby. Somehow a couple of turnstiles continuously break down and the customer starts to suspect the issue is quality related due to the recurrence of issues in these specific turnstiles. Using data from the access control system it was determined that the turnstiles in question saw far more traffic than other turnstiles. As a result of the information contained in this data, the customer was able to make changes to spread the load out more evenly between all turnstiles.
The most prolific supplier of raw data in this industry is video surveillance, supplying not only the viewable image itself, but a nearly limitless amount of data culled from each element of the scene. This is called metadata, and all analytics rely upon the generation of metadata to apply algorithms in order to draw conclusions.
The conclusion may be detecting a face, recognizing a face, determining the direction or speed of an object, characterizing the difference between objects, determining color, and so on. But all of the analytic algorithms rely on the generation of metadata. It is this generation of metadata that not only enables the video analytics, but will serve as a foundation for monetization of intelligence and insight for business and operational insight.
As users seek to refine and analyze the data, the tools they use have to be intelligent enough to do that and in a meaningful way. For example, this can mean going beyond telling a retailer that peak hours are between noon and 1 p.m., but also draw conclusions around weather and traffic. As the goals of the user are more fully defined, these intelligent systems that process and analyze data from video streams can also make the decision as to whether reduce staff levels or bring on additional resources.
HVAC sensors combined with video sensors can create complementary data to measure occupancy patterns and usage of areas to determine when to turn lighting on or off or when to operate heating and cooling. That way a building that will soon be unoccupied is not continuously being heated at a higher level, which in turn creates savings.
Big Picture Issues With Big Data
The cloud environment that supplies the massive processing power and storage capacities required by big data presents both opportunities and concerns for end-users and systems integrators navigating this new environment. As the scale of stored information grows, so too does the stakes of a cybersecurity breach. End users involved with big data need the assurance that the third-party repository storing their data is secure and that privacy of their information is a primary focus.
For example this includes making sure the data is partitioned in way so that it stays in a certain region or country, as privacy laws vary by region and certain countries in the world have very strict privacy laws on data.
As the reliance on and use of big data continues, integrators will find themselves in the position of dealing with it in some way. Although some may take on the role of helping to manage big data, most will likely find the biggest value comes in understanding it and helping customers select a system that can both generate and, through its network, manage the data.
Dealing with a cloud-based enterprise also brings security together with IT, so integrators should either have the personnel in place to deal with this, such as hiring someone with a data-focused background, or be able and willing to interface with someone beyond the security director within the business they are servicing. And there are also additional costs to consider with implementing and maintaining these systems.
Big data management presents opportunities for integrators to become part of the value chain, even if they are not the one handling the primary function of collecting, transmitting, and analyzing data. But it also requires the understanding that this is a paradigm shift from selling hardware to selling a service with various features and capabilities.
Who among an integrator’s clientele is most likely to be a big data user? Currently, enterprises with high-risk profiles such as casinos, airports, energy-generating facilities, and pharmaceutical companies can be counted among those businesses that would benefit from in-depth analysis of their data.
The day will come, however, when big data will have an impact beyond key verticals and for those who have gotten in ahead of curve, the rewards will be greatest.