A tutorial to be held as part of ISWC 2008.
Organizer: Dr. Elena Gaura, Reader in Pervasive Computing
Director of Cogent Computing Applied Research Centre, Coventry University,
Faculty of Engineering and Computing, Tel: +44-(0)24-76888909, e.gaura@coventry.ac.uk
Co-tutors: Dr. James Brusey, Senior Lecturer, Cogent Computing Applied Research Centre, Coventry University, Faculty of Engineering and Computing, Tel: +44-(0)24-76887687, j.brusey@coventry.ac.uk
Goal: This is a practical tutorial accompanied by a live demonstration of a working, application-integrated body sensing and actuation wireless system. The application scenario supporting the tutorial is that of manned bomb disposal missions. The tutorial focus is on the generic design processes for deployable, wearable, wireless sensing and actuation systems.
Abstract:
Wireless sensor networks (WSNs) offer immense potential for performing detailed multi-parameter measurements in a variety of applications. When coupled with actuation, this technology could become even more powerful.
WSNs have been a buoyant area of study for over a decade. Two research strands have been mainly followed: theoretical and large scale application feasibility studies on the one hand, and small scale, practical deployments on the other, with a large gap lying between these two strands. (For example, theoretical studies of wireless sensor networks have tended to involve highly sophisticated approaches to gathering data from distributed nodes, while practical implementations in industry have tended towards much simpler, point-to-point approaches that minimize complexity.) Given that such theoretical approaches to WSN design treat, by and large, all target applications as aspects of the same problem, subsequent proposed designs are complex and sophisticated. When they form the basis for practical deployments, these designs have to be considerably stripped down for two main reasons:
particular application requirements impose constraints unable to be met by the theoretically driven cumbersome and overweight designs;
much of the technology and techniques taken as given in theoretical designs are neither sufficiently mature nor sufficiently well characterized to be applied in viable, real world deployments.
It is clear that the approach of stripping down complex systems to achieve practical deployment is both expensive and lengthy. Application led designs rarely need the amount of complexity available at a theoretical level. Instead, concerns for robustness, data integrity, ease of use, long-life, reliability, and maintainability take over as primary design concerns. Practical deployment design processes can be considerably sped up by starting with simpler systems that are more focused on smaller sets of target applications. The design concerns specific to that application set will naturally lead the design process in terms of selecting what off-the-shelf hardware and software can be used, and what bespoke components need to be developed to satisfy the application as a whole.
This tutorial will support the exposition of design techniques and design choices by focusing on an example from within the area of Embedded Body Sensor Networks. The application is that of remote health monitoring of human subjects during safety-critical missions. The embedded body sensor network is neither large nor widely distributed but there are a number of fundamental requirements (such as, the size of the nodes, wearability of the instrumentation, robustness, reliability and fault-tolerance, etc) that dictate the majority of the design and implementation choices.
WSN technology has not yet reached wide scale adoption. However, pursuing application driven design processes will enable the development of industrially strong WSN systems which will, in turn, increase confidence in the technology and contribute to its adoption in near future.
The audience does not necessarily need to be well-versed in the domain of wireless sensing.
The tutorial will be of interest to a wide range of conference participants: designers and developers of real-life sensor networks applications for the Body Sensing domain; developers of integrated miniature, embedded technology for smart textiles; evaluators of wearable technologies; end-users of body-related embedded technology (clinicians, physiotherapists, bio-mechanics engineers, etc)
Introducing the Wireless Sensing domain (30 min, EG)
Wireless Sensor Networks: matching application requirements with available technology (30 minutes, EG)
Case study: a wearable mission critical body sensor network (40 minutes, JB)
System purpose and requirements description
Application stake holders and end users
Demonstration of body sensor network system capability; detailed temperature mapping, posture assessment, heart rate monitoring, modelling of thermal sensation in network, remote observation point functionality and features, remote and in-network actuation of cooling for comfort
The sensing component: (30 minutes, EG)
What would we want to sense and what sensors would we use to sense those parameters?
The computational component: (20 minutes, JB)
How do we go from sensor data to useful information?
The communication component: (20 minutes, EG)
Where is the information needed and how can it be transferred there?
The actuation component: (20 minutes, JB)
How do we close the loop — deciding on how to act on the basis of the information gathered?
Putting it all together into an application (30 minutes, EG / JB)
How do we wrap the components together to meet the application requirements?
Evaluating wearable systems (30 min, EG/JB)
Reader in Pervasive Computing
Director of Cogent Computing Applied Research Centre
Faculty of Engineering and Computing
Coventry University, UK, CV1 5FB
Expertise: Advanced Measurement Systems; Design and Deployment of Wireless Sensor Networks; Intelligent Sensors; MEMS Sensors; Artificial Intelligence integration.
Elena received her BSc in Electrical Engineering from the Technical University of Cluj Napoca, Romania in 1989, following it with an MSc two years later. In 1991, she joined the same institution on a professional basis. In 1996 she came to the UK, initially as a Research Assistant at Brunel University, Uxbridge.
In 1998 she joined Coventry University, in order to pursue her PhD, researching the integration of Artificial Intelligence (in particular Neural Networks) and MEMS sensors to produce enhanced performance Microsystems. By the time her PhD was awarded in 2000, she was serving as a Senior Lecturer in Computer Science. With the establishment of the Cogent Computing Applied Research Centre in 2006, Elena was appointed as the inaugural director.
Over the course of her career, Elena has accrued a sturdy academic reputation. She has over 100 refereed publications, and is a frequent organiser of symposia at Nanotech (NSTI), the world's largest nanotechnology conference. She is a member of many nationwide Microsystems advisory bodies and, since 2002, a member of the EPSRC College of Peers. She has recently become the Co-Chair of the UK Wireless Intelligent Sensing Interest Group (WiSIG).
She remains highly active in the research field, exploring issues of Microsensors Integration and Large Scale Wireless Networks, deploying MEMS technology to develop new techniques for integrating sensors, actuators and networking functions, with the ultimate aim of designing autonomous systems capable of large scale field sensing. In addition, she is engaged in examining the potential of AI techniques to solve design problems, particularly drawing on wireless sensing and AI integration.
Senior Lecturer in Systems Engineering
Senior Research Fellow of Cogent Computing Applied Research Centre
Faculty of Engineering and Computing
Coventry University, UK
James Brusey received his BApSc with distinction in Computer
Science from the Royal Melbourne Institute of Technology (now
RMIT University), Melbourne, Australia, in 1996. In 2002, the
same institution awarded him his PhD in Artificial Intelligence
for a thesis entitled Learning Behaviours For Robot
Soccer
, which also won the Australian Computer Science
Association's award for Best Computer Science PhD Thesis for
Australia and New Zealand in 2004. He has augmented his research
reputation with wide ranging professional experience in both
industry and academia. His early experiences in the Swinburne
Institute of Technology twined both threads together: as a
systems programmer he maintained the in-house software and
mainframe, whilst also lecturing on x86 Assembly Language.
In 1989, he began working for AXA Australia as a systems programmer. The reliable and robust systems he developed earned his team two company awards for excellence. He left AXA in 1996, but within two years had returned to perform legacy mainframe maintenance and provide training for staff in diagnostic methods and mainframe assembly languages. This experience embellished his industrial training acumen acquired during his time as a technical consultant with Storage Technology Australia (StorageTek).
In October 2002 he came to the UK as a Senior Research Associate in Cambridge University's Institute of Manufacturing, initially on a grant from MIT's Auto-ID Centre. At Cambridge he acquired significant hands-on experience of working with RFID, constructing a realistic RFID demonstration, employing industrial robots. This formed a key deliverable for the Auto-ID Centre project. The Centre subsequently became EPCglobal Inc, and is responsible for the EPCglobal set of standards and specifications governing the use of RFID in the retail sector. These standards are founded on the principle of an Electronic Product Code (EPC), which uniquely identifies everything from forklift trucks to fizzy drinks cans. He joined the Cogent team as a Senior Lecturer in 2007.
His current research strands are: Robust system design approaches for Wireless Sensor Networks, reactive approaches for actuation, and Bayesian approaches to state estimation for Body Sensor Networks.