Ubiquitous Context Aware Rule-based Environment


Ubi-CARE is an intelligent framework developed by the EML2 group in the home-healthcare domain. Its purpose is to collect heterogeneous raw data coming from different biomedical sensors associated to different patients and to analyze them in a context-aware way. According to Ubiquitous Computing and Context-Awareness paradigms, Ubi-CARE performs a wide data gathering and a logic elaboration of such data in order to determine anomalies in vital signs of the patients and to notify alerts with different priority levels to appropriate medical operators, according to their actual availability.

Ubi-CARE seems to be very useful in the homecare domain, where patients haven’t got the same monitoring systems as when they are hospitalized.

Our framework architecture integrates an EM component with an advanced IT infrastructure.  The former is represented by a RFID system (Radio Frequency IDentification) that gathers and sends raw data coming from physical sensors; the RFID system employs specific passive, low-cost, multi-purpose tags . The latter manages raw data without requiring the user intervention and regardless the patient location.

Let’s imagine a patient monitored at home: an increasing blood pressure that exceeds a given threshold will generate an “anomaly verification” event. This event will be propagated throughout the system until it reaches the proper medical operator

From an architectural point of view, our framework (depicted in the figure below) relies on a multi-agent system. The software agents implemented perform:

-    reasoning (thanks to a rule-based system)

-    formal communication (thanks to an ontological model shared by all the agents)

Sensors and databases are the physical and virtual sources respectively.

The final results (anomaly and alert notification, patient actual localization, patient actual vital signs…) are notified in a proper graphical way.