Aspinity develops an analog machine learning processor that enhances battery-operated, always-on sensing devices by improving energy efficiency and extending battery life by ten times. This technology enables precise event detection and classification in applications such as IoT, smart home, and wearable health monitoring, while minimizing power consumption from irrelevant data processing.
Funding
$16.5M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
+2Founders
Product
Problem
Many battery-operated, always-on sensing devices face limitations in energy efficiency, resulting in short battery life due to continuous data processing, even when most data is irrelevant. Traditional digital signal processing consumes significant power, hindering the development of truly always-on, low-power IoT, smart home, and automotive security applications.
Solution
Aspinity offers an analog machine learning (AnalogML) processor designed to enhance the performance and energy efficiency of always-on sensing devices. The processor leverages analog processing to minimize power consumption while maintaining the accuracy of machine learning for event detection, classification, and prevention. By processing data in the analog domain, the system significantly reduces the amount of irrelevant data that reaches the power-hungry digital components, resulting in a substantial increase in battery life. This technology enables devices to remain in a continuous monitoring state with minimal power draw, allowing for quicker response times and more comprehensive data capture.
Target Audience
The primary target audience includes manufacturers of IoT devices, smart home systems, automotive OEMs, and developers of wearable health monitoring solutions seeking to enhance the battery life and performance of their always-on sensing applications.
Features
- AnalogML core enabling near-zero power always-on system operation (typically <100μW).
- AML100 processor with a library of automotive surveillance algorithms, leveraging sensor fusion from acoustic, piezoelectric, and radar inputs.
- AML200 processor scaling to 22nm, supporting up to 125,000 parameters for complex machine learning applications such as RF and image processing.
- Software programmability for adapting to various event detection and classification tasks.
- Automotive security solutions for parked vehicle monitoring, including acoustic-only triggers for detecting events like door handle jiggling or glass breaks.
- IoT Enablement Platform for rapid deployment and evaluation of always-on sensing applications.