Introduction
As global populations age, the number of seniors living independently is rising, bringing a corresponding increase in the risk of unobserved medical emergencies such as falls, strokes, or heart attacks. Traditional health monitoring solutions often suffer from high latency—recording data for later analysis rather than immediate action—or network inefficiency, where raw data streams overwhelm mobile connections. The iCare system addresses these gaps by combining wireless body sensors with smartphone-based local processing to provide a real-time, context-aware health platform that doubles as a "living assistant."
System Architecture
The iCare system is built on a four-tier architecture designed to be non-intrusive yet highly responsive. This structure ensures that data is not just collected, but acted upon immediately at the point of care.
- Sensors and Devices: Wearable Bluetooth-enabled sensors, including ECG leads, accelerometers, and blood pressure monitors, collect physiological data from the user.
- Smartphone (The Hub): The smartphone acts as the central processing unit. Unlike systems that merely relay data, iCare performs local real-time processing. It compares incoming vitals against pre-set thresholds. To mitigate false alarms, an anomaly must be detected in two consecutive cycles before an alarm is triggered, giving the user a window to manually cancel if the reading was caused by interference.
- Server-Side Intelligence: A central server hosts the Personal Health Information System (PHIS) and a Medical Guidance database. This tier stores long-term history and allows doctors to remotely update a patient’s monitoring thresholds via SMS as their condition evolves.
- The Social Loop: Uniquely, iCare integrates family and friends into the care process. The system provides tiered access, allowing social circles to view health history and cooperate with medical professionals.
The "Living Assistant" Concept
Beyond vital signs, iCare is designed to assist with the daily challenges of aging. The system includes a Climate Reminder, which pulls data from weather servers to advise users on appropriate clothing, and a Medicine Reminder to assist those with memory loss. This holistic approach transforms the device from a medical monitor into a proactive daily companion.
Implementation and Results
The prototype was implemented using Java ME for the mobile application and JSP for the web-based server components. Testing confirmed that the system could successfully manage real-time data and facilitate remote management by medical staff.
Key results from the prototype phase include:
- Integrated Emergency Response: The system demonstrated the ability to send emergency SMS messages containing the user's precise GPS location to both medical centers and family members.
- Dynamic Thresholding: Doctors successfully pushed new monitoring parameters to the patient's phone remotely, allowing for personalized care without requiring an in-person visit.
- Knowledge Inquiry: A medical database was established that returns health advice based on keywords, with results categorized by "confidence levels" to ensure reliability.
Limitations and Future Work
While the prototype proved the feasibility of the iCare architecture, several areas for improvement remain. The current user interface requires a more accessible design for users with declining vision or motor skills. Additionally, as the medical knowledge base expands, the authors plan to implement more advanced AI optimization to prevent database bottlenecks. Future iterations will also focus on improving indoor location accuracy and scaling the model to support regional emergency services, ensuring that help is dispatched from the nearest possible location.
Summary updated