Warning: This post may contain triggering and or sensitive material. Gun violence and mass shootings are some of the sensitive topics mentioned within this blog. If uncomfortable with these topics or anything similar, I kindly suggest you do not read on.
A few days ago, on March 22nd, I sent the text that no one wants to send, “Are you ok?”, eventually followed by, “Were you near the shooting?”. One of my good friends attends the University of Colorado Boulder, and was within a mile radius of King Soopers, a grocery store in Boulder and the most recent location of another massacre in America. Thankfully she was safe, but the incident in which a patron opened fire in a supermarket, resulted in 10 deaths. Unfortunately, I’ve sent more texts checking on loved ones after receiving a breaking news alert then I could ever imagine, and am sure this is a reality for many of us in current day America.
When entertaining ideas for my blog this week, this incident came to mind. I instantly dismissed the thought of blogging on anything related shootings and gun violence because it was heavy and can easily get political. As time went on I came back to that thought and realized that I shouldn’t shy away from this topic because it matters.
That being said while writing on this topic, my goal is to maintain objectivity and analyze this issue from a lens of pragmatic technological implementation, mainly focusing on increasing public safety through digital technology so we can be better equipped and ultimately decrease deaths and mass shootings within the US.
“Technology cannot end depravity or violence, but it can limit the evil a person can inflict on others.”Jeremy Shane, CNN
Acoustic Gunshot Detection
The use of technology to locate the origin of gun fire has been used by militaries dating back to World War I. This technology has developed greatly since then, and has applications in today’s society that look quite different (InfoWatch). Gunshot detection systems and sensors today are being used to help decrease response times and notify security, first responders, and authorities within seconds of an incident occurring. The technology mainly takes form of sensitive microphones and sensors, and sometimes include cameras. Most gunshot detectors enlist artificial intelligence and machine learning to help increase the accuracy of detection and reduce the probability of false detections. By using a library of millions of sounds, these systems often boast impressive accuracy.
An innovative leader within this space is a company that goes by the apt name, ShotSpotter. Their patented gunshot detection technology uses artificial intelligence, algorithms, and deep learning to detect, locate, classify, and respond to incidents. They have reviewed over 15 million incidents to date, and use a machine classifier to compare these sounds to a large database of known gunfire. This classification, along with other “impulsive community sounds” helps the AI to determine if the incident is gunfire or not. If the incident is classified as “likely gunfire”, it is then sent to acoustic experts in their Incident Review Center to be analyzed and either published to police, or dismissed if deemed not gunfire. ShotSpotter’s process is illustrated in the diagram below.
This entire process, from initial gunfire to alert takes less than 60 seconds, and results of ShotSpotter implementations are impressive, as seen below.
ShotSpotter’s technology can help to mitigate tragedies and ultimately help save lives.
“In July, after a sniper shot 12 officers in Dallas, police armed a small robot with a bomb and detonated it next to the shooter—killing him, and ending an hours-long standoff. It was the first time police in the U.S. had ever used a robot to kill a suspect.”William Brennan, The Atlantic
Different from these technological infrastructure approaches taken by ShotSpotter, Knightscope is taking a different, more futuristic approach. Formed in response to the tragic Sandy Hook shootings and Boston Marathon bombings, RoboCop is becoming a reality sooner than expected with their K1, K3, K5, and K7 models. These robots, pictured above, are fully autonomous security robot. Knightscope’s models are currently servicing malls, casinos, hospitals, offices, airports, and more, and can be found patrolling North Central Texas College’s campus and the NBA’s Sacramento King’s home court. The K5 ASR enlists sensors and high definition cameras to patrol an area that has been defined using GPS. This technology mainly scans for anomalies within a specified area, but Knightscope developed their first prototype for a Visible Weapon Detection System in 2019, which uses machine learning to spot firearms and alert authorities if one has been recognized. A main differentiator of the ability of physical robots, is their potential in counteracting the threat and fighting back, potentially firing at the threat or using a nonlethal force to incapacitate or distract the force. Although this is not currently a capability of Knightscope’s robots, these capabilities are being explored. The use of this technology has shown impressive results within a public park as seen below, and this idea of using robots as a “force multiplier” is becoming increasingly popular. The results of implementing Knightscope security in a public park are shown below.
Implementation and What’s Next
Increasing public safety and digitalizing infrastructure to help decrease possible harm done by shootings and crime is vital. Technologies such as the Knightscope could even fight back in the future, firing at a potential threat or incapacitating them. With this power of course, comes major responsibility, that Missy Cummings, Head of Robotics at Duke University does not believe these robots are ready for (The Atlantic). Her fears include the crowded nature of places that these mass shootings usually occur, which presents extreme technical challenges. These capabilities would require much more computing power than the robots currently possess. Many of these implementations are also extremely costly for high stakes places such as schools and supermarkets. Using the Boulder shooting as an example in a grocery store, over the last 20 years supermarkets revenue growth in the US has been flat, which is not promising for undertaking large technological transformations to increase public safety. With all this in mind these are high stakes situations, and it all comes down to the possibility of saving lives. Through the implementation of technology, although it may not be able to fully stop evil, we can hopefully minimize the damage done by these attacks and be able to send a few less “are you ok?” texts to loved ones.