nir-vana innovation network

The Czech tech innovative company was founded in 2014 by a team of engineers with extensive experience in computer networking, cloud storage systems, machine learning and embedded development. Their main focus is the creation of standalone solutions for the security and public safety markets. The company has developed a device that detects specific types of unwanted acoustic events such as human scream, glass breakage, bomb explosion, gunshot or urban noise and enables an immediate response by a security or safety team to a dangerous situation by sending them alerts within 5 seconds. It is intended to protect public areas as well as the private security sector and can be used outdoors as well as indoors. Typically, the technology is used in smart cities (placed on busy streets), banks, healthcare facilities, educational institutions, airports, public and private spaces. The device is driven by a machine learning algorithm and contains a deep neural network trained on a huge database of sound records which can distinguish normal events from dangerous ones. The algorithm recognises acoustic events in real time and it is not necessary to send any data outside of the device which means that the solution is 100% GDPR compliant. The solution can be integrated with 3rd party software and hardware. The Czech company has recently integrated their software into the security cameras of a world-leading manufacturer. Thus, customers using these cameras do not need any special equipment and can integrate the software inside the camera and equip the camera with an external microphone. The SW solution comes with a very intuitive user application that is precompiled as a native mobile or desktop version and can be customized. The sound event detector is already used in many cities in Europe, Brazil and the USA. Additional features of the sound event detector: • Coverage design - enables to visualize how many devices are needed to cover a specific area and estimate the best position of the device • Noise measurement - the device is able to measure the level of noise in its surroundings • Localisation of the sound event using at least 3 devices