FAQs

Home / FAQs

FAQs

AmberBox Gunshot Detection uses tri-factor authentication to deliver rapid gunshot detection, without false alarms.

 

In addition to infrared muzzle flash and percussion sensing (the characteristic noise of a bullet leaving the chamber and the crack of a supersonic bullet), the system also utilizes bespoke machine learning sound analysis. Each detector incorporates this patented machine learning analysis, comparing sounds detected against a database of over 3,000 gunshot samples, stored on every detector. This ensures fast, accurate and reliable detection at the detector head, without communicating any real time audio data across the network, guaranteeing privacy. Importantly, by utilizing our tri-factor authentication, AmberBox Gunshot Detection does not require line of sight to trigger, an issue faced by existing infrared and percussion only approaches.

 

Upon activation, automatic alerts are made to police via our E911 certified telecommunications platform and notification to building security representatives. Threat alerts are issued by call, SMS and push messaging to building security representatives.

 

Instant response technology removes human error and the bystander effect, allowing authorities to respond to threats faster and more accurately. Automatically protect an area—no user intervention required.

Amberbox Gunshot Detection workflow

AmberBox can integrate with a range of security features and provides security personnel an easy-to-use interface to quickly respond to the situation.

The device has been tested up to 300ft and has never had a false activation or a non-activation. The device has been calibrated to 60-90 ft spacings to ensure total accuracy in its detection capabilities. As a rule of thumb, each device will cover a 5000-7500 sqft area.

Yes, AmberBox can detect through walls. The range of detection can be affected by wall thickness.

Yes, the system can detect subsonic ammunition.

 

Silenced weapons have a very different audio and infra-red signature compared to unsilenced weapons. We are currently teaching the detector to sense for silenced models. Once we can reliably detect this and have sufficient silenced weapon data to maintain our near-zero false alarm rate, we will push the update to the detectors as part of the service. However, it is important to note that silent weapons account for less than 0.1% of gun violence crimes.

AmberBox detectors can be fitted on ceilings or walls, typically at 3ft spacings away from any other electronic or lighting fixtures.

No. What is unique about AmberBox is that all the analysis is done at the detector head. The processor in the device undertakes a comparison function on the audio and infrared data input, activating only upon the unique signatures of a gunshot. The wireless communication protocol physically limits the bandwidth preventing real-time audio transmission. Furthermore, the network receives its own unique AES encryption key and data that is transmitted between devices always receives a new temporary encryption key every time data is hopped, ensuring security.

The AmberBox set up begins with a gateway being connected to the internet. This can be done via a GSM  or Ethernet connection. One gateway can coordinate up to 250 detectors.

Amberbox Gunshot Detection network

Once the gateway is connected to the internet, the detectors  self-initialize via a IEEE 802.15.4 protocol that meshes the detectors together. This allows the sensors to quickly pass information back to the gateway where the activation can begin its external communication to police dispatch, AmberBox’s Security Operations Center and other security operatives. Therefore, no other connection (like WiFi) is required to individual detectors.

AmberBox has achieved a near-zero false alarm rate. The detection algorithms utilize machine learning to determine what is and is not a gunshot based on a massive set of data that is continually updated. Detectors receive regular over the air updates to ensure that this false positive rejection is always optimal.