Aizip Intelligent Time-Series

A foundation model for time-series data has been developed to detect anomalies and provide alarms for preventive maintenance.

The model can run on ARM Cortex M3 and up processors, or comparable RISC-V, NPU, DSP, and FPGA devices.

This domain-general model can be used off-the-shelf with a quick registration process. No retraining or fine-tuning is required.

This model provides accurate projection of electricity usage by leveraging historical data, which can be used to optimize energy load.

This efficient time-series model can run on ARM Cortex M0 and up processors, or comparable RISC-V, NPU, DSP, and FPGA devices.

Saving energy usage is a key effort to achieve a sustainable world. This model series can be expected to be widely used.

This model provides detection of multiple classes of human activities, such as standing, walking, going upstairs, and going downstairs.

This class of efficient models can be as small as 8kB, and run on ARM Cortex M0 and up processors, or comparable RISC-V, NPU, DSP, and FPGA devices.

This model can be adjusted for other pattern detection and classification, such as livestock activity and bridge vibration.

Accurate fall detection can be achieved with this model when used in conjunction with wearable sensors.

This compact, efficient, and robust model can run on ARM Cortex M0 and up processors, or comparable RISC-V, NPU, DSP, and FPGA devices.

This model provides very low false positive rate, which is a key challenge for products in this application.

These are a few select examples. More available upon request.