Arcturus Networks has said it has demonstrated an audio signal processing software subsystem to detect alarming audio noises such as screams, gunshot, smoke alarm, Glass-break, Domestic carbon monoxide alarm, T3 Evacuation Signal (ISO / ANSI /NFPA), Dog bark, Baby cry, Electronic personal safety alarm, Assault whistle, Hand clap and any Undeclared audio event.
The subsystem is designed to operate both in connected or as a standalone system. The system samples the ambient audio and check for any change in audio amplitude. For any change in audio levels, the software subsystem starts processing audio signals features such as temporal, frequency, pattern, amplitude and other characteristics are used to create an acoustic fingerprint which is then compared with existing profiles to determine a match. The system also transmit the audio data to a remote processing server based on the trigger.
With the support of M2M networking, it is possible to tweak detection profiles, minimize false positives, increase accuracy and extend the practical application of the system. The solution is intended to coexist with full duplex audio applications such a two-way voice communication, engaging and disengaging as required. The system provides a unique differentiator for service providers looking to maximize the value of an in-unit two-way voice communication system and remote monitoring infrastructure, according to Arcturus.
“The combination of an audio sensor with our M2M and VoIP middleware, means that security monitoring applications have an early warning system that gives them explicit information about the profile of an emergency situation, before emergency services are even notified,” said David Steele, product manager, Arcturus. “Being able to provide this information, in advance, may significantly improve the preparedness of a response and the overall safety of all involved.”
Pricing and Availability
General availability of the solution is scheduled in Q2 2014.