(Press Release - March 13, 2026) – As Taiwan enters a super-aged society, demand for smart healthcare and long-term care technologies continues to grow. AI startup Deepfens AI Inc. has recently launched the market’s first multimodal edge AI fall detection system, targeting high-risk areas such as bathrooms and restrooms in hospitals and long-term care facilities. Through AI sensing and behavioral analysis technologies, the system provides real-time safety monitoring. This solution partners with Avalue Technology Inc. (TPEx: 3479.TWO), a provider specializing in industrial computing solutions and medical embedded systems, leveraging Avalue’s medical-grade edge computing platform to deliver a comprehensive smart healthcare solution.
According to Taiwan’s National Development Council, the proportion of people aged 65 and above has exceeded 20%, marking Taiwan’s official transition into a super-aged society. With the increasing elderly population, hospitals and long-term care institutions are facing growing pressure from staff shortages and safety management challenges. Among these issues, falls are one of the most common accidents among elderly individuals. Bathrooms and restrooms, in particular, are high-risk environments but are difficult to monitor with traditional surveillance systems due to privacy concerns, making them a key challenge for smart care technologies. Deepfens AI noted that the core goal of smart care is to detect safety incidents in real time while protecting user privacy and promptly notifying caregivers. Unlike traditional camera surveillance or wearable devices, the system utilizes millimeter-wave radar non-contact sensing technology to identify human posture and motion changes in real time. This allows the system to detect falls or abnormal behaviors without the use of cameras, ensuring both privacy and safety.
The system also integrates voice interaction capabilities. When abnormal behavior is detected, the system can actively communicate with the user to confirm their condition, helping reduce false alarms. By combining multiple AI modalities—including body movement, positional changes, and environmental information—the system enhances overall detection accuracy. From a system architecture perspective, Deepfens AI adopts an Edge AI design, enabling AI inference to be processed locally on the device without uploading sensitive data to the cloud. This approach ensures both data privacy and low-latency response. Over time, the system can also analyze behavioral patterns such as daily activity levels, nighttime mobility frequency, and prolonged inactivity, providing healthcare teams with valuable insights for patient care.
On the hardware side, Avalue Technology collaborates with Deepfens AI to integrate the AI sensing system with medical-grade embedded computers and edge computing platforms, enabling hospitals and long-term care facilities to rapidly deploy smart care solutions. Through this partnership, a one-stop solution is provided—from sensing devices and AI analytics to care management platforms—supporting applications such as fall detection in patient rooms and bathrooms, in-hospital safety management, post-discharge home monitoring, and activity tracking for chronic disease patients.
The showcase will feature a compact AI Boc PC MAB-T660D powered by Intel® 14th Gen Raptor Lake Refresh processors, built on the Intel® H610E platform and supporting up to 64GB DDR5 memory for high-performance edge computing. The system also provides PCIe Gen4 x16 expansion for integrating AI accelerator cards. In addition, Avalue will present the HPS-GNRU4A high-performance server, powered by Intel® Xeon® 6 processors, capable of supporting four dual-width GPUs and equipped with PCIe Gen5 high-bandwidth interfaces. This design offers flexible and high-speed computing expansion, enabling efficient data transmission and centralized management while significantly enhancing the real-time response capabilities and operational efficiency.
Deepfens AI stated that in the future, smart care solutions will gradually expand into home healthcare and smart city applications. By leveraging AI sensing and behavioral analytics, these systems aim to safeguard the safety of elderly individuals and patients without disrupting daily life, while also providing healthcare systems with richer behavioral health data to advance patient-centered smart healthcare models. The solution will be showcased at the 2026 Smart City Summit & Expo from March 17 to 20 at the AI Zone Pavilion. Industry partners and professionals are warmly invited to visit Booth Q1023 to experience firsthand how multimodal Edge AI technologies, combined with medical-grade computing platforms, are powering the next generation of smart healthcare and long-term care solutions.