Keynote Speeches
Key Note Address 1

Pervasive Sensing and Computing
Srikanta (Sri) P. Kumar
Program Manager, DARPA, USA

Pervasive Sensing and Computing (PSC) will penetrate many walks of human life transforming the way we instrument and control our surroundings. This talk will review recent technology developments in the area of integrated pervasive sensing and computing, and address research challenges ahead. Progress and trends in defense and commercial applications will be presented. The talk will cover recent advances in micro-sensor technology including hardware platforms, networking and collaborative computing methods, as well as middleware and PSC applications. Key accomplishments from related DARPA programs will be presented. Challenges in transforming sensor data to knowledge supporting decision making will be covered. Potential approaches to dramatically reduce device costs and form factor through chip-scale integration of processing, communications, and micro and nano-sensors (MEMS and NEMS) will be discussed. Algorithmic and software challenges for building high-functionality context-aware systems that are also robust, reliable, and secure will be presented. Potential directions in application of cognitive information processing techniques including machine reasoning and learning to achieve rapid adaptation (of PSC systems) to changes in contexts and environments will be discussed. The talk will also present applications, near and far-term, along with associated research challenges.
Key Note Address 2

Beyond Global Communications: The Active World
Roy Campbell
The University of Illinois, Urbana Champaign

The confluence of pervasive computing, anywhere/anytime access to information resources and scalable computing enables the construction of smart environments or Active Spaces. In such a Space, a spectrum of computation and communication devices seamlessly augment human thought and activity with digital information, processing, and analysis to provide an observed or imagined world that is automated and enhanced by the behavioral context of its users. The power of such a computer infrastructure has three contributing factors; the translation of information to and from physical properties, the computers and their ability to transform data, and the cooperative computational environment that results from embedding these devices in a network. This computational environment or "Active World" is the likely long-term benefit of the current information technology revolution. Several major projects have shown the benefits of considering pervasive computing environments within an infrastructure, constructed from computing elements that interact to form active or smart spaces, and managed by a software control system or meta-operating system to provide integrity and consistency. As a case study, our experimental system, Gaia, creates a pervasive computing environment that encompasses multiple rooms of our new building: the Siebel Center. Tasks like tours, exhibitions, seminars, lectures, or meetings are supported by coordinated distributed applications and both tasks and their contents may be programmed. Mobile users within the building are tracked with location sensors and may create sessions involving different tasks which they then may migrate, suspend, or resume as they move from room to room. Despite recent advances, many challenges remain. Integrating the various services, components, applications, and entities into a programmable COTS infrastructure enables context sensitive applications that allow users to interact seamlessly with a combination of physical and computer facilities. Such an infrastructure of smart devices, rooms, and buildings raises the question of how to manage, program, automate, and formalize these heterogeneous sources, sinks, repositories, and processors of data. The organization, management, and programmability of physical devices and information activities in a pervasive computing environment is key to enabling diverse, autonomic, digital habitats such as university campuses, office buildings, scientific labs, and museums. However, the promise of pervasive computing cannot be realized without cost-effective and efficient mechanisms, policies, and tools to organize, manage, operate, repair, program, and evaluate systems built from pervasive computing components. Human tasks, human factors and pervasive system infrastructure interact in complex ways and methodologies need to be devised to explore and measure these interactions. In particular, a pervasive environment would need to enable opportunistic collaboration, facilitate social interaction, and support teaching and learning. This talk will explore the benefits of an Active World, the barriers to its deployment, and the research challenges that lie ahead.

The Accomodation site has been updated.
Please make your reservations by
Feb 7, 2005.



Sponsored by
The University of Texas at Arlington, IEEE Computer Society,
IEEE Technical Committee on Computer Communications (TCCC),
IEEE Technical Committee on Parallel Processing (TCPP)


Visited by individuals(counter is not counting multiple access from one site)