Automobiles use 100+ sensors to continuously monitor what’s happening in and around the vehicle while driving and while parked. But sensing, collecting, and processing all of this data consumes too much energy using current technologies and always-on system architectures. This ultimately places time limits on always-on monitoring when vehicles are unattended.
The Aspinity AML100 analog machine learning processor combines the event detection accuracy of machine learning with near-zero power analog event detection and wake-up, delivering integrated and after-market automotive activity detection solutions that consume just 10’s of μAs to detect:
- Other activity
enables the lowest power always-on AI dashcam solution, significantly extending the amount of time that dashcams can operate in park mode.
Additionally, Aspinity’s machine learning impact detection algorithm ignores non-impact sounds or vibrations so there are fewer recordings for the dashcam to store and the car owner to review.
- AML100 consumes <20µA when in always-listening mode
- Camera records only events related to the vehicle and ignores the rest
- Detects general impacts and/or classifies specific impacts such as a door open or a window glass break
- Evaluation kit available
The AML100 enables a near-zero power, high performance automotive solution that can be used for integrated vehicle security and monitoring, vehicle access, or other always-on applications.
- Monitor parked vehicles without draining the battery
- Sensor modules integrated into multiple car panels
- AML100 sends data to CPU only when touch event is detected
- Aspinity sensor fusion ML/AI algorithm discriminates between different types of impact events to trigger various vehicle responses (alarm, unlock, etc.)
for more information about our integrated automotive solution.