Atmosic's partnership with AONDevices will reduce product power consumption by reducing false detect wakeups, RF transmissions, and cloud time.

Atmosic Technologies, an innovator of ultra-low-power wireless IoT platforms such as the award-winning ATM33/e, today announced its strategic partnership with AONDevices, a pioneer in low-power application-specific edge Artificial Intelligence (AI) processors including the AON1120. The collaboration brings together low-power AI capabilities and ultra-low-power wireless technology, for innovative AIoT and wearable applications.

Atmosic's ATM Series extreme low power wireless SoC includes integrated power management and energy harvesting capabilities, the RF radios to support standards-based protocols such as Bluetooth Low Energy and 802.15.4 protocols such as Thread and Matter, memory, and the application processor along with a suite of peripherals including GPIO, ADCs, serial interfaces, keyboard scan matrices and more. AONDevices brings a suite of AI technology and their AON series of AI processors that are incredibly accurate even in the presence of high background noise or reverb conditions, offering reliable voice command recognition, sound and contextual awareness, and multi-sensor fusion. When used in voice control or situations requiring context detection, their technology supports speaker independence in always-on and multi-word detection scenarios with adaptive voice activity detection.

Atmosic has set a golden standard in IoT sustainability by reducing the power consumption of RF radios, and by leveraging harvested energy to power sensors, tags, remotes, and more. Atmosic's solution enables developers to create greener products by reducing reliance on batteries that can leach toxic chemicals into landfills and water tables when incorrectly disposed of. The partnership with AONDevices brings further power reductions by cutting energy used when sensors 'false detect'.

"When a sensor incorrectly triggers, additional power is consumed by the device itself and by the supporting cloud services as they process the invalid information," said Mouna Elkhatib, CEO of AONDevices. "Using AI to verify that a trigger was indeed valid before the downstream energy consuming events occur, helps reduce radio transmissions & re-attempts, and the processing/storage of data by cloud services."

Mike Fortin, SVP of Sales from Atmosic added, "In security applications, for example, if a sensor can accurately diagnose the sound of breaking glass, as opposed to sounds played on a television, then radios can remain in their sleep state extending the energy reserves in a sensor for when it is needed."

AI models can be created in only a matter of hours and both Atmosic and AONDevices offer comprehensive development kits that accelerate time-to-market for customers looking to create products for applications including headsets, wearables, game controllers, toys, or in vehicles, smart home appliances, smart buildings, or industrial applications.