AWS Public Sector Blog

The Future of Policing: Detect, Locate, and Alert on Gunfire in Under a Minute

For every gun-related homicide, there are 100 gunfire incidents that go unreported. But what if every incident was reported, and police departments could adjust and deploy resources based on real-time analytics and data delivery?

ShotSpotter delivers real-time gunshot notifications to law enforcement so they can dispatch to the precise location of the gunshot, engage with the community, look for evidence and occasionally help victims and make arrests. Their goal is to help drive down illegal gun use.

The ShotSpotter solution is an excellent example of the AWS Cloud enabling next generation technology for the public safety community. For a small business, ShotSpotter needed to cover a broad range of technologies spanning sensor hardware design and embedded software development, classification of gunshots through machine learning, development of algorithms for precisely locating gunshots in a challenging acoustic environment, sensor network management, API development, and front-end technologies for delivering gunshot notifications and analysis.

“When we started out on the development of our new front-end technologies around nine months ago, one of our primary goals was to host the backend in a commercial cloud, rather than continuing to invest in expansion of our private IT environment,” said Paul Ames, SVP of Products and Technology at ShotSpotter.

ShotSpotter started out with a core set of AWS services, including: Amazon Elastic Compute Cloud (Amazon EC2), Amazon Virtual Private Cloud (Amazon VPC), and Amazon Route53 for on-premises and commercial cloud communication. As they continue to work with AWS, they will look to adopt Auto Scaling and Elastic Load Balancing.

“Running a private IT environment, you spend a lot of cycles defending against intrusions and denial of service. There’s no glory in this work, and so anything that we can do to lessen the overhead on our staff while at the same time improving the quality of defenses is welcome,” said Paul.

The increasing acceptance of the value of ShotSpotter to public safety agencies is driving large increases in the number of square miles of coverage the company provides.  This kind of growth is beginning to become a challenge. “Provisioning new backend infrastructure as we grow is clearly a lot easier with a commercial cloud,” said Paul.

In addition to horizontal scaling, ShotSpotter has two very critical times a year when vertical scaling becomes a real issue: New Year’s Eve and July 4th. Its outdoor gunshot sensors are acoustic, and fireworks present a real challenge in terms of classification, especially since there is an increase in incidents detected by two orders of magnitude. What they need is to be able to scale up compute leading up to these celebrations and then scale back down again when everyone settles down.

“For an on-premises cloud, it’s ultimately cost prohibitive to own and maintain infrastructure sized for peak periods,” said Paul. “At this point in our evolution, with just the first phase of customer endpoints delivered via AWS, we’re dealing with a load that’s a lot more predictable and isn’t significantly peaky. We’re looking forward to deploying our detection, location and classification algorithms to AWS to solve our celebration challenge.”

“ShotSpotter’s solution demonstrates the value of making data available in real time to officers in the field,” said Michael Wagers, Global Public Safety Lead, Amazon Web Services. “It is our hope that by arming law enforcement with next generation tools and resources, like ShotSpotter, they will be able to respond to situations faster and ultimately keep our communities safer.”

Follow us on Twitter @AWS_Gov with #futureofpolicing. And watch our Periscope interviews with leading police chiefs from IACP 2016.