9.00-9.20 am.
Title:
Overview of UWB-enabled Sentient Computing (UWB-SC) Research Programme
Speaker:
Assoc Prof Chen-Khong Tham, Institute for Infocomm Research (I2R), Singapore

9.20-9.50 am.
Title:
Impulse Radio UWB Real Time Localization System
Speaker:
Assoc Prof Choi-Look Law, Nanyang Technological University, Singapore
Abstract:
This session first presents the results of UWB ranging experiments conducted in an indoor environment consisting of line of sight (LOS) and heavily blocked non line of sight (NLOS) scenarios. The UWB measurement campaign with over 1,600 points was conducted in a typical indoor office environment covering an area size of 25m by 25m and with range of up to 30m in an open hall to assess the sensitivity of ranging errors to parameter settings (such as threshold setting and sampling rate) and receiver architecture. Results show that centimeter to decimeter level accuracy is obtained with coherent and non-coherent receiver. Even in NLOS scenarios, decimeter level accuracy is obtained for 60% of the measurement points. This is followed by the results of a real time localization system based on time difference of arrival.

9.50-10.20 am.
Title: Efficient ADL Recognition using Wireless Sensor Networks
Speaker: Assoc Prof Chen-Khong Tham, Dept of Electrical & Computer Engineering, National University of Singapore & Institute for Infocomm Research (I2R), Singapore
Abstract:
Activity of Daily Living (ADL) monitoring is important in order to determine the well being of elderly persons in their home settings. Some important questions are, “Is the elderly person able to eat properly on his own? Is he able to walk steadily in the home?” In this talk, we present some results of our work on human activity recognition: (i) eating activity recognition using wireless sensors and Dynamic Bayesian Networks (DBN), and (ii) human walking using wireless sensors and the extended Kalman Filter. In the case of (i), the DBN algorithm reduces the complexity of determining states. Initial results are quite promising and point to a general algorithmic approach that uses multiple modalities of sensors for gathering data, detects activity primitives and stores detected activity primitives as micro-context for future use. In the case of (ii), adaptive sensor selection is used to select sensor nodes with lowest uncertainty to minimize sensing interference. Both techniques are also used to facilitate resource management in resource-constrained wireless sensor nodes by powering down sensor nodes which are not necessary for achieving accurate recognition at a given time.


10.20-11.00 am.
Tea Break and View Posters and Displays / Demonstrations

11.00-11.30 am.
Title:  Smart Space enabled by Sentient Computing
Speaker: Assoc Prof Wee Ser, Nanyang Technological University, Singapore
Abstract:
Imagine a physical space filled with “experts” whose role is to provide personalized services to a certain target group of people in that space. Clearly, such a space is capable of taking care of the well being of the target group of people in that space. The question is, can we do it without involving real human experts as the later are expensive and can be put to better use. This talk will present a project that attempts to provide such a smart space by equipping the space with sensors and a signal processing system, that perform the functions of “eyes”, “ears”, and “brain”. The example used for the space in this talk is a home, the target group of people in that space consists of two “elderly” human subjects, and the personalized service provided by the system is the monitoring of the activities of daily living. The talk will present the design of our system and some technical challenges encountered by the project. Some results obtained by the project will also be shared. Specifically, some demonstration video clips will be shown on the detection, recognition, and tracking of two human subjects, foreground segmentation, 3D reconstruction and behavior recognition of the human subjects, recognition of human behaviors relating to meal-taking, and detection and recognition of audio sounds at the scene. This project is part of the A*STAR Thematic Strategic Research Program on “Ultra Wide Band-enabled Sentient Computing”.


11.30-12.00 noon
Title: A Service-oriented Context-aware IT infrastructure for Pervasive Healthcare
Speaker: Assoc Prof Hung-Keng Pung, Dept of Computer Science, National University of Singapore
Abstract:
The worldwide need for a better healthcare support among the growing aging population faces a number of challenges, including rising medical cost, inadequate number of medical doctors and healthcare professionals, and higher incidence of misdiagnosis. One promising solution to these problems is the development of a service oriented context aware IT infrastructure for pervasive health/wellness-care. It allows accurate and timely delivery of health/medical information among patients, doctors, and healthcare workers through widespread deployment of wireless sensor networks and mobile devices. These networks can be used to support health care related services such as: patient monitoring, location-based emergency response system, anomalous activity detection, and pervasive access to medical data. These pervasive healthcare services will greatly improve the quality of service of medical professionals as well as widen the scope and accelerate the delivery of health/wellness services to greater number of people. How to delivery these services across heterogeneous social and health/wellness-care domains is a big challenge. In this talk, we will present a potential solution known as Healthcare-COALITION.

12.00-12.30 pm.
Discussion

There will be poster presentations and a couple of demos on the same day of workshop.