In today’s fast-moving social, mobile and digital world, not being plugged into the constant deluge of potentially relevant information is a scary concept. Most of us can’t make it an hour without checking our emails, text messages, friend status and social connections on Facebook, Twitter, LinkedIn or Instagram. A very small percentage of this information that we receive is of actual interest to us, with the majority being noise and junk. However, we feel obligated to check it anyway, spending the time to weed out the important ones to get to what matters to us.
Weeding out the important data and creating an actionable process is not only an issue on an individual level, but also in government – especially the Public Safety function. Numerous information feeds are available to public safety departments – video surveillance streams, analytics, geo-alarms, CBRNE sensors, social media monitoring, event management, incoming 911 calls; the list evolves as quickly as new technologies become available. Public Safety leaders are actively looking for ways of managing this high-volume, time sensitive problem: how do you filter down to the important data, to create meaningful, shared situational awareness that triggers an actionable response?
The oft-used term situational awareness has evolved along with the technologies that deliver it. Standalone systems – like video surveillance – provided initial visibility into an organization’s environment. Then greater integration of systems and devices connected additional behaviors into a common operating picture for response purposes. The entry of situational management systems (like PSIM) further connected data collection devices with standard operating procedures to produce enhanced response. Today’s available toolset – or Situational Awareness 4.0 – further leverages Big Data and Predictive Analytics into the mix, offering greater awareness and valuable insights not previously available with its predecessors. By pulling the combination of Big Data and Predictive Analytics through a PSIM, public safety can easily manage and monitor not only social media activities but also GIS, 911 calls, CAD data, LPR, Video monitoring and analytic data, incident management data and many others. Let’s examine how we may arrive at this promising combination of technologies by walking through key situation awareness tools available today.
Maximizing the Value of Your VMS to Leverage Influx of Video Data to Monitor and Prevent Public Safety Threats.
Public safety agencies - at all levels and functions - have deployed sophisticated video management systems (VMS) to inform their operations, but have come to find that monitoring thousands of live video footage with limited resources is unreasonable. Watching hours and hours of video recordings and locating incidents after the fact is very time-consuming and quickly exhausts valuable resources. As if this wasn’t enough of a headache for public safety officials and their IT counterparts, now we see the pending proliferation of body-worn cameras throughout law enforcement agencies. With the national spotlight on recent police shootings and with over 36 states with active body camera legislation, law enforcement agencies through the US have been readying their personnel and IT departments and actively piloting body cam technology.
Already, IT departments have identified the need to deal with the tremendous amount of data generated by this and other video surveillance feeds (e.g. traffic cams, private sector video). Multiple terabytes of data can be produced on a daily basis. The sheer quantity of this data, associated mandated retention policies, CJIS security compliance, and the complexity of managing and distribute video amongst multiple levels of law enforcement operations across multiple types of devices (e.g. phones, tablets) demands a thoughtful solution. Our firm, SDI, has worked with the evolving category of Enterprise Video Content Management systems (EVCM) to fit this growing public sector need. EVCM is a software platform that allows for the ability to store and share video data irrespective of its source and to manage that data in response to a complex matrix of access and use requirements. Proven in its adoption in the private sector, EVCM’s cloud-based architecture allows organizations to centralize, manage, and deliver large-scale video content online while providing more flexible, accessible, and configurable storage for long-term evidence purposes. By adding EVCM into the mix of a public safety agency’s video environment will allow more data and content to be accurately analyzed for even greater situational awareness.
Advancement of such video analytics has helped agencies tremendously, but issues still remain. Requiring operators to bypass false alarms or provide instruction to self-learning analytics is time consuming to the agencies. More often, operators at the monitoring centers are overwhelmed with false positive notifications and overlooked useful information. In addition to wasting valuable resources and time focusing on the low-value notifications, one of the biggest dangers of too many false alarms is that when real events occur, alarms and notifications may not be treated seriously and doesn’t received appropriate attention. Next generation analytics have moved from a post-event, forensics tool to a preventive tool – more accurately identifying precursor behaviors to trigger proactive intervention. Also, as the resolution of surveillance cameras continues to evolve, facial recognition analytics that incorporate behavioral algorithms will prove to be a more effective and accurate solution.
Deploying advanced technology that extracts the maximum value out of video also plays out in regional response centers, including fusion centers. When the DHS funded “premature” fusion centers between 2003 and 2007, massaging the mountains of intelligence gathered into meaningful action was a significant issue. Analyst at these and other centers continue to face these issues every day, with more and more data coming their direction to analyze, making quick decision making harder and harder. The last several years has brought the introduction of artificial intelligence-based analytics and other more sophisticated video analytic algorithms, aiding in the effort of identifying points of value and incidents more accurately and in a timely manner. Additionally, PSIM providers are implementing this sophistication into their applications, allowing the tagging of events and archiving them for easy access for immediate and future evaluation of prioritized content. More proactively, responding units to a tagged incident can receive automatic e-mail notifications and snapshots of alarms and incidents so they know what to expect when they arrive on the scene. Such rapid and dynamic information exchange is providing responders with the capability of retrieving live footage around the scene for further situational awareness.
Graduating to a Situational Awareness platform (aka PSIM)
At SDI, we often see differing functionality expectations between a well-integrated VMS and PSIM technology. A category of software created by middleware developers, PSIM is designed to integrate multiple unconnected security applications and devices and control them through one comprehensive user interface. Traditional PSIM – when deployed correctly – gives tremendous power to operators by delivering multiple sources of intelligence into a common operating picture. PSIM is a perfect fit for most sophisticated public and private operation centers. It is PSIM’s core functionality to consolidate numerous sub-systems together to create a cohesive common operating platform to present actionable picture to the decision makers. From aviation to 911 centers, transportations to financial institutions, PSIM can monitor, manage and consolidate all the notifications under its platform. Multiple end users that range from security, operations, facilities and maintenance can share a common understanding of an incident.
However, in order for PSIM to live up to its fullest potential, it requires significant investment in its initial deployment, as well as on-going consideration when an organization chooses to implement new technologies. For instance, most public safety agencies are using, in one form or another, social media to monitor for actionable content. These types of systems must be incorporated into an agency’s PSIM investment, or again it will again exacerbate public safety operators’ ability to identify actionable intelligence.
For those organizations not ready to take on PSIM levels of investment, we see an emergence of ‘PSIM lite’ in Digital Video Management Systems (DVMS) space. DVMS vendors are increasingly offering built-in integration with core security systems such as access controls, alarms and video verification, providing functionality approaching that of PSIM. The process and resource requirements demanded by PSIM are cost-justified by organizations that are consolidating multiple command and control centers, and are prepared to map their Standard Operating Procedures into the system’s logic. Stick with VMS or a PSIM lite product when looking for a more limited functional scope, such as managing alarms, integrating with access control, and recording and disseminating video.
Enter Big Data and Predictive Analytics
Recently, we have been seen Predictive Analytics and Big Data often used in same sentence in articles, online blog posts and conversations. It is not surprising to see this correlation between Big Data and Predictive Analytics because together they offer a perfect solution for analyzing massive data sets and providing tools to make accurate decision. Before we begin to describe the relationship between Predictive Analytics and Big Data and how – together with PSIM – they are being used to predict the future outcome for the public safety, let’s talk about each concept separately and how they complement each other:
Predictive analytics is the practice of mining complex information from existing data sets with the purpose of determining common patterns, and using these patterns to create models to predict future outcomes to help businesses to make intelligent decisions. Predictive analytic models do not ”magically” tell you what is going to happen in the near future; however, it predicts what might happen in the near future with a degree of reliable outcome. This outcome can also include what-if scenarios, leading to incident management, disaster recovery and risk management recommendations.
Even though Big Data is a familiar concept for the private sector, it is fairly new for Public Safety entities such as Fire, Police, 911 centers, non-emergency call centers, Health Departments and Aviation. Big Data becomes relevant to our conversation as the information being collected by public safety organizations is both structured and unstructured (variety), in large amounts (volume), and has properties of exponential growth (velocity). Examining video content and drawing correlations to other data types like texts, photos, social media content in Twitter, Facebook and Instagram feeds - than you are dealing with Big Data. Big Data cannot be analyzed using traditional data processing tools within a tolerable timeframe; it requires a specialized skillset and sophisticated tools. But analyzing Big Data is only half the challenge. The other half is generating results that are usable and actionable to the first responders, and this is where predictive analytics and then SAMS will be helpful.
Using predictive analytics’ sophisticated algorithms and software to analyze the Big Data will help first responders and other public safety agencies monitor, measure, and predict criminal activities. Working examples today include predicting when and where criminal activity is most likely to take place: in order to prevent criminal activities, agencies can assign police resources where they will have the most impact. Building on that example, analytics can track criminal activities and spot new and emerging crime trends, to create crime maps and hot spots, and improve early detection of threats and potential terrorist activities. Predictive analytics, fueled by public safety Big Data, can be used for both manmade emergencies and natural disasters. Predicting the aftershocks of earthquake or hurricane and taking appropriate measures to stop criminal activities; combining atmospheric dispersion plume modeling to base hot-spot policing strategies; coupling gunshot detection with historical crime data to get accurate crime type patterns are a few examples of how first responders can use Predictive analyses and Big Data.
Large-scale Connectivity of the Future
Going back to the importance of incorporating new technology into a public safety organization’s PSIM investment – this must include these Big Data and predictive analytics applications. This powerful technology collaboration will allow public safety managers to incorporate these heightened insights into their response protocols and SOPs. It will allow public safety operators to manage millions of messages and notifications in real-time and create actionable response with minimum effort. The power of Big Data and predictive analytics – delivered through a PSIM platform – can take a public safety organization from just trying to keep up, to staying one step ahead.
Developing this advanced public safety environment is certainly a complex undertaking. Integrators with public safety operations expertise and working knowledge of all these uncoupled technologies can provide agencies with the orchestration needed to deliver advanced situational awareness. The right system integrator will optimize each individual technology, create connectivity and data flow between the solutions, and deliver productive insights most needed by public safety user community.
About the Author:
Yilmaz Halac is Director of Client Delivery for SDI. He previously served as Managing Deputy Director City of Chicago Office of Emergency Management and Communications (OEMC), overseeing technology for Chicago Police (CPD), Chicago Fire (CFD), 311 (non-emergency) and OEMC - 911. Halac was a member of Public Safety Technology Consortium, a group which consists of CFD, CPD, OEMC, Health Department and Aviation. He was responsible for the research, planning and development of new technologies for all the public safety agencies. Halac is well-credentialed in large-scale, event-based security management, having worked on the planning for the recent NATO Summit in Chicago in 2012, as well as serving as a member of the Critical Infrastructure Protection Subcommittee along with the US Secret Service, FBI, DHS and other high profile supporting agencies. Prior to this major event, Mr. Halac also played a significant role in other major events including 2008 Presidential election in Chicago. Mr. Halac has presented at numerous national and international technology conferences. He has several DHS certifications and recognitions from various government organizations.
