Aspinity, a pioneer in ultra-low-power analogue machine learning (AnalogueML) processors, has launched its Voice-First Evaluation Kit (EVK2), a complete hardware/software development kit.

The EVK2 enables customers to integrate analogue machine learning and data compression into battery-operated voice-enabled devices, such as hearables/wearables, smart speakers and smart TV remotes, facilitating significant power savings but without compromising on system accuracy.

The kit features the latest generation of Aspinity’s Reconfigurable Analog Modular Processor (RAMP) chip, the world’s first implementation of a compact, ultra-low power analogueML. The RAMP chip has introduced a new architectural approach to system design that improves battery life in edge devices.

In contrast to alternative always-listening system architectures, which digitize all sound data, relevant or not, before wake word analysis, the RAMP chip uses near-zero power to analyse raw, unstructured analogue microphone data at the start of the signal chain to determine if voice is present prior to triggering the wake word engine.

Since up to 90% of the sound data captured within a day is not voice, the RAMP chip’s analyse-first approach is intended to minimise the power-on time of the analogue-to-digital converter (ADC) and wake word engine (WWE), increasing battery life by up to 10x.

The RAMP chip is also the first analogue voice wake up solution to continuously collect and compress (into ~2kB of memory) the 500ms of sound prior to the wake word (pre-roll) that is required by most WWEs in order to accurately determine that a command has been spoken.

The Voice-First EVK2 features:

  • The latest-generation RAMP chip, along with voice activity detection and pre-roll collection, compression, and reconstruction algorithms, a one-source solution for high-accuracy, low-power always-on voice wake up
  • Audio test files for quick start-up, and a live audio testing option that uses a high-performance MEMS microphone from Infineon for flexible testing
  • Integration with the popular high-performance STM32H743ZI MCU from STMicroelectronics which allows the testing of analog voice activity detection with or without preroll collection and delivery to a third-party WWE

“The EVK2 jumpstarts the integration of the RAMP chip into power-efficient voice-enabled devices,” said Tom Doyle, CEO and founder, Aspinity. “For the first time, device designers can realise all of the benefits of analogueML and analogue compression - 10x power savings without a reduction in wake word detection accuracy - for their next generation of voice enabled devices.”