Alongside supervision of PhD students, teaching, and administration my interests are with wireless sensor network (WSN) systems design for real-life applications. I have a strong belief that contributions at systems level are needed for the sensor networks technologies to become a widespread reality and be adopted commercially in a variety of application domains.
Functional sub-components of such systems are presently being developed by a variety of research groups, approaching the field from a number of viewpoints (a wealth of routing protocols have been published for example, not to speak of the variety of architectures that have been proposed). However, many a time, such functional components are developed somewhat in isolation from either the higher level application requirements or the lower level physical end of the system (the sensors and actuators). Most importantly, there is seldom sufficient thought to the real-life deployment of systems containing the developed sub-components.
With a design space as diverse as that exhibited by generic WSN systems, many notable successes in WSN design and implementation have come from work that aims to cater for particular subsets or individual constraints that the computing, communication, or sensing technology imposes (see, for example, the constraint of limited energy resources). Nonetheless, achieving robust deployments, effective, trustworthy and timely information extraction from collected data and short design-to-implementation cycles implies adopting a higher level approach to the design of WSN systems. The seamless integration of plug-and-play micro-machined sensors with the computational platforms that enable the networking of sensor nodes in large numbers, for example, is particularly challenging. It is safe to say that any other plug-and-play feature raises numerous technical, intellectual and creative challenges.
My personal vision for sensor networks is that of scalable, easy to construct, easy to deploy and use systems.
As an engineer, I am a strong believer in experimentation. Hence, adopting an real-life, application driven view over the WSN design space, coupled with the production of cyclic, physical instantiations of the WSN design process, is, for me, most meaningful. Such an approach not only caters for, but brings to the core of the design the user
and the application itself.
Whilst no two WSN applications have identical requirements, to ensure success and continuity, drive towards generic design is needed to allow for rapid solutions to be developed. This is particularly important when working in a dynamic research environment with a variety of industrial partners and application domains.
The key generic aspect which allows for re-use of WSN functional and conceptual components is an encompassing framework to enable (and, further on, at implementation stage, support) pluggable elements as follows:
However none of the above is possible without the development and use of rapid prototyping and debugging tools.
Pursuing the research above counts heavily on an ability to form prime teams of engineers, computer scientists and application domain specialists. Personally, I delight in working within multi-disciplinary teams, hence, I am hopeful for a bright research future for myself and those I work with, within the new enabling paradigms of networked sensing.