April 2017 Digital Edition
March 2017 Digital Edition
Feb. 2017 Digital Edition
January 2017 Digital Edition
Nov/Dec 2016 Digital Edition
Oct 2016 Digital Edition
Preventing the Boston Bombings with a ‘predictive security system’
The disastrous results of the Boston Marathon bombings are testimony to the ineffectiveness of current security systems to predict and help avert devastating attacks. Although Boston’s security infrastructure was quite effective at post-disaster forensics, enabling the identification and neutralization of two suspects only days after the bombings, the community still had to suffer hundreds of injuries, loss of lives and hundreds of millions of dollars in collateral damage.
Tragedies such as the Boston bombings are driving a movement towards the development of “Predictive Security Systems” designed to anticipate and avert the kind of devastation that we witnessed during the Boston Marathon bombings.
Predictive security systems operate in a manner that simulate the presence of hundreds of highly trained human operators analyzing real-time video footage collected by hundreds of surveillance cameras. Video data leading to the capture of the Boston bombing suspects was available prior to the attack, but it was simply impractical to expect human operators to be watching in real-time the right video streams to see the evolution of an attack threat profile. This is where “Predictive Security Systems” play a major role because they utilize technology that simulates the cognitive properties of the human thought process to anticipate threats based upon relentless analysis of real-time video footage.
The technology is based upon “Symbolic Cognitive Architectures” and “Inference Process Algebras,” which have already been utilized in robotic systems, such as NASA’s deep space and planetary probes. The technology is quite sophisticated and when properly utilized embeds a cognitive feature into security systems allowing these systems to alert first responders to the possibility of an impending attack.
Specific to the Boston bombing, a cognitive security system would have instantly detected the following four observations and actions through well-known video analytic processes:
- Two individuals wearing backpacks walking together with a gait suggesting the backpacks are heavy and probably not filled with clothing.
- The same two individuals separating near the finish line on Boylston Street by about 630 feet.
- The same two individuals almost simultaneously taking off their backpacks and placing them on the sidewalk.
- Both individuals leaving the area without their backpacks.
Each of these observations and actions taken alone may not be a cause for alarm, but taken together they represent a unique threat profile suggesting the possibility of an attack. A trained human operator lucky enough to observe this video data in real time would probably have come to the conclusion that an attack was imminent, but given the vast volumes of video data the chances of an operator being available are next to impossible. Predictive security systems with a cognitive capability would instantly detect these observations and actions and come to the same conclusion as a trained human operator, alerting first responders within seconds of the suspects dropping their backpacks and walking away.
A cognitive system would also continue monitoring the location of the suspects in real-time as the area around the backpacks was cleared; and even though the backpacks exploded about 15 minutes after they were dropped, there would have been time to clear the area of people, preventing injuries and death. Finally, the location of the suspects would have been tracked in real-time enabling them to be apprehended and captured, preventing the massive multi-million dollar manhunt that ensued after the bombings.
The Boston Marathon tragedy has driven home the realization that today’s physical security systems must be upgraded with a cognitive capability that enables systems to predict impending attacks, rather than just analyze attacks after devastation has occurred.
James Ionson, Ph.D., is CEO of Cognovizum Inc., which is pioneering the emerging field of predictive cognitive security systems. He can be reached at: