Low-power motion recognition with algorithms running directly on Leantegra beacons. Smart processing for less power consumption.
Fall detection, walking, running, inactivity and other scenarios.
Machine Learning for Human Motion
- Training and classification algorithms suitable for ARM Cortex-M0 MCU and tiny amounts of RAM.
- Much less transfers via Bluetooth – no raw accelerometer data, just the recognition events.
- Supported natively by Leantegra WiBeat and WiRange devices, as well as the mobile SDK.
- Utility app for Android – data collection and ML training purposes.
Better RTLS with Motion Data
- Better accuracy and smoothing for RTLS – with motion data fusion in Leantegra RTLS engine.
- Better power consumption for RTLS hardware – less wireless transmissions during the inactivity periods.
- Better installation options for RTLS anchors and tags – less dependencies on NLOS and coverage limitations.