Consumer electronics, smart home, and industrial application developers are demanding greater efficiency and easier control for their audio-based interfaces; however, this paradigm is challenged by the requirement to analyze ALL of the audio data in search of relevant information, as well as the corresponding power that is needed to achieve this. As a result, Aspinity has focused on audio as a lead application area for its revolutionary analog processing solution. Customers can realize an order-of-magnitude system power reduction while maintaining "always-on" operation with complete analysis of the audio data received from the microphone. Aspinity’s RAMP can be programmed to detect virtually any "sound signature" within a broad set of audio data, and subsequently trigger a wake-up or provide specific data for post processing.
Voice control is in high demand because it promises an intuitive user interface that will work in situations where other user interfaces are inconvenient. To respond to voice commands at any time, devices must always listen, but this is the primary challenge—to achieve sufficient battery life while still accurately detecting voice commands in the presence of background noise. Crucial to meeting this challenge is voice detection. Since less than 10% of audio typically contains speech, there is great potential to enhance battery life by minimizing power consumption during non-speech durations. Unlike other voice-detection solutions—which require the operation of an analog-to-digital converter, memory, etc.—Aspinity’s solution minimizes power by only powering the microphone and the Aspinity ASP. Integrating Aspinity’s ASP into a voice control signal chain affords an approximately 10x reduction in voice detection power, which extends to a similar reduction in average power to detect an activating keyword.
In industrial systems, machinery wears down or needs maintenance on a regular yet unpredictable timeframe. In these systems, monitoring the vibration characteristics offers a glimpse into early troubles and allows preventative maintenance. One challenge in vibration monitoring is how to handle the large quantities of vibration data that are generated by vibration sensors distributed throughout a structure or plant. Vibration data are captured at a moderately high frequency (up to tens of kHz), and thus generate a significant amount of data for an embedded sensing device. This abundance of data requires more powerful processors and greater power consumption. Vibration monitoring devices digitize all the data and then perform an FFT on the data. Peaks in the FFT spectra indicate the vibrational modes of the monitored structure, and the health of the structure can then be inferred from the relationship between the vibrational modes, and a vibrational spectral alarm can be triggered for maintenance action.
The desired vibration information is contained within a relatively small number of data points (the modes), so by extracting information earlier in the signal chain with Aspinity’s RAMP architecture—prior to digitizing the signal—the necessary resources for an embedded vibration monitoring device can be significantly reduced. Extracting these modes with Aspinity's analog core processor provides a 100x reduction in the data that must be digitized and handled by the processor: from 2048 data points for an FFT frame to 10 pairs of frequency and magnitude datum. While data reduction is a key benefit to vibration monitoring applications, Aspinity's analog processing solution will provide additional benefits to power, size, and costs of the entire system.
Capacitive touch input propelled the adoption of smart phones; now new features, such as 3D gestures and proximity-based activation are driving new applications and requirements for capacitive-sensing technologies. The critical component for touchscreen “intelligence” is the touch controller, which measures the touch-sensor array (typically capacitive), digitizes this measurement, and then analyzes the sensor data in a microcontroller to extract commands or actions for the system. To support next-generation user interfaces—including three-dimensional gestures (swipes, circles) and wake-on-proximity—tomorrow’s touch controllers must provide higher signal-to-noise ratio (SNR) and sensitivity. Unfortunately, today’s touch controllers oversample sensor data to obtain the necessary precision, however this oversampling does not achieve the precision required for next-generation interfaces with ever decreasing power and size budgets.
Aspinity offers an alternative solution where the touch controller’s analysis tasks are implemented in its RAMP architecture so that each measurement is precise, thereby alleviating the need for oversampling, and achieving a higher precision-per-joule. Aspinity’s analog processing solution reduces the power consumption of the touch system by more than 10x for a comparable SNR. Additionally, implementing the touch controller in analog processing will remove the need for the digital processing circuitry, which will enable a smaller size.
Interest in continuous health tracking has driven the demand for wearables. But achieving low-power operation (i.e. long battery life) for the diverse range of health sensors has proven challenging. Aspinity’s technology has been proven for applications such as heart-rate monitoring via both ECG sensing and PPG sensing, in both cases achieving ultra-low power consumption.