Falls are a health killer for the elderly. If the fall can be reported and rescued as soon as possible, the harm can be minimized. There are many types of technologies used in fall detection. Currently, the accuracy of image posture recognition technology is relatively high, up to 95%. After an incident occurs, care givers can confirm the authenticity of the incident through video playback, and report back. AI algorithm can be continuously optimized and improve the accuracy. AUOCare Sensetek uses a skeleton (stick figure) method to display body movements in the image, which can avoid privacy concerns. It is suitable for use in living rooms or public spaces (such as indoor corridors or stairwells/entrances). This type of camera with fall detection can replace traditional surveillance equipment and can achieve the effect of active alarm. More importantly, it can be combined with back-end rescue services so that care givers in the field can assist nearby.

- The AI ​​algorithm calculates and recognizes postures based on the joint points on the skeleton, reducing the impact of body shape and clothing.
- Using edge computing, only the action characteristic data (skeleton image) is transmitted
- Provide fall event notification and risk analysis