創新及科技解決方案

解決方案編號

S-0017

解決方案名稱

Falling Alert System for Elderly Care-- Optical Monitoring System (OMS)

解決方案描述

OMS proposes its camera solutions as a better alternative, which does not require elderlies using any wearable technologies, and could instantly detect a falling motion which alerts the person-in-charge. The AI solution components are facial recognition that differentiates the identity of all elderlies inside the center, 3D key-point recognition that detects the motion of falling, and object recognition that identifies objects that might cause the fall. The solution components are all under 1 system.

應用領域

城市管理
發展
房屋
環境
基礎設施
人口
社會福利

使用的技術

機器學習
人工智能
深度學習
數據分析

使用例子

Falling is considered the most common source of injury amongst the elderly population. Approximately 30%-50% of falls cause minor injuries such as bruises and abrasions, and around 10% result in major injuries, including fractures and brain injury-related deaths. Elderly centers have been actively trying to prevent the falls, but accidents do happen and an immediate response could avoid the situation to worsen. The existing solution to this problem is the ‘safety bell’ that requires elderlies to always wear it on their necks and push the button whenever there is an emergency- such as they have fallen. Although the device could instantly contact nearby hospitals and inform police stations, it is largely dependent on Elderlies pushing the button, hence when they are unconscious or thinking they could manage by themselves, will often delay the best time to receive medical treatments and worsening the situation. 

The falling alert aspect uses facial recognition that identifies the elderly, and key-point recognition that recognizes the body key-points such as joints, limbs etc hence knowing when there are irregular body movements causing a fall. When a fall is detected, an instant alert will be sent to the end-user through an online dashboard (APP/Web) allowing elderlies to receive immediate care preventing the situation to further worsen.

公司／機構名稱

Dayta AI Limited

電郵地址

patricktu@dayta.ai

電話號碼

98628809

網址

http://dayta.ai

地址

716C, Enterprise Place, 5 Science Park West Avenue, HKSTP

如政府部門對創新及科技解決方案進行概念驗證或驗證技術請聯繫智慧政府創新實驗室。