Division of IT Convergence Engineering POSTECH’s U-Health Monitoring and Support Smart Home for the Elderly Improve quality of life & reduce healthcare costs Enhance comfort and safety of elderly at home Reduce stress and burden to family members Design & testing of interface circuits with sensor elements Progress in intelligent sensing, monitoring, and semantic decision making Synergistic application of ICT, BT and NT for health and social well-being of humanity, especially elderly persons Demo Smart Home at POSTECH General Research Framework Smoke Detection Autonomic Computing Part Importance & Motivation Allow elderly to live independently and safely Significant reduction in health-care costs Many aspects of elderly well-being and safety can be automated Early detection of health problem early treatment less expensive and better health Improve chronic and geriatric care at home Bring healthcare to remote locations & in poor countries through convergence application of ICT, NT and BT Service Provider Autonomic System Data Network Healthcare Providers ( Control System ) Medical Wireless Sensors Network Sensor Monitor Analyze Plan Execute Knowledge Home Gateway Networked Appliances Cellular Network Smart Home Service Smart Home Services Contract “Services To The Home” Contract U-Health Service SpO2 Motion Sensor RFIDTag Appliance Control Heart Beats Sensor Presence SpO 2 Sensor Open Services Framework Knowledge Base AI Rules Reasoning Decision Making Core OS Platform Physical Environment Sensors Actuators MonitoringAnalyze Execute Smart Home + Humans Plan Smart Home Components Health care part Hospital / doctor Specialized organization Remote diagnosis Autonomic computing part Information filtering / aggregation Situation / context modeling Intelligence reasoning Decision making Home networking part Information gathering Service discovery Appliance discovery Sensing part Actuator Home control unit Home automation Smart Home Bio sensor Environment sensor Light Path for Obstacles avoidance ECG Sensor Wireless Sensor Base Conclusions Research Challenges Environment discovery Addressing and routing Self-organisation and Self-healing Networks composition and mobility Virtualization Context modeling Self -detection algorithms Self-diagnosis algorithms Intelligent sensing and monitoring Learning Sensors Actuators Home networks Autonomic system Health care Compatibility with existing techniques and healthcare models. Interdisciplinary collaboration Social/Societal implications $ $ $ $ $ Compatibility with low-cost standardized manufacturing Intelligent sensors & actuators design Ultra low power design of integrated circuits and systems Best cost-performance reliability Autonomic decision-making Home Network Part Ontology model for U-health smart home 1.Requirements a.Semantically Rich Knowledge Base: Capture concepts and relationships b.Dynamically Updateable Knowledge Base: Enhance with new information during lifecycle c.Context awareness: Situation awareness in smart home & identify specific contexts. d.Support semantic reasoning: Infer new facts and update Knowledge Base. 2.SHOM (Smart Home Ontology Model) a.Using OWL (Web Ontology Language) to define classes and relations between them b.OWL-DL (Description Logic) based on SHOIN Description Logic. OWL-DL ensure decidability c.Concepts related to Smart Home Network, Appliances, Humans. 3.Decision Making a.Data gathering through medical and environment sensors. b.Data aggregation, fusion and filtering c. Inferred information using First-order Engine WBAN ( Wireless Body Area Network) Information-based sensor scheduling to improve energy efficiency and low latency Motivation Specific disease Determine relevant parameters of symptoms. High-level information Key relations among these symptoms. Determine best body sensor(s) for specific parameters of symptoms. Quantify sensing and communication operations for diagnosis. Propose cooperative diagnosis models for different body sensors. Approach An Information-based probabilistic relation model A Cost function over the energy expenditure A correlation model between utility gain and energy loss Internet Pulse ECG Temperature SpO 2 Accelerometer Respiratory rate Network Coordinator ZigBee GPRS caregiver Emergency Medical Server EEG Hui Wang 1, Hyeok-soo Choi 2, Nazim Agoulmine 1,3, M. Jamal Deen 1,4 and James Won-Ki Hong 1 1 ITCE, POSTECH, Pohang, South Korea 2 Computer Science & Engineering, POSTECH, Pohang, South Korea 3 Computer Science, University of Evry Val d’Essonne, France 4 ECE Department, McMaster University, Hamilton, Ontario, Canada Algorithm System Architecture Statistical Analysis Knowledge Base Decision Making Coordinator subject to Sensor 3 Sensor 2 Sensor 1 Data Actuators Start Initialization Sensor selection Wait for information Update knowledge Finish Knowledge good enough? Yes No How the world evolves? context Situation assessment Revise goals Goals Generate/revise decision rules (SHOM) Sensor tasking/action Information utility Compute initial knowledge Send information query