Advanced Technology Forum

Tuesday, May 4, 7:00 PM

MEASUREMENT, MODELING, AND COMMUNICATION IN SENSOR NETWORKS

Jan Smid

Loyola University Maryland
Columbia Graduate Center
8890 McGaw Road

This talk is about practical and theoretical design and implementation issues of integrated sensor, database, and decision systems.

In the future, distributed systems for monitoring and modeling will be based on flexible communication among their components and their users. Systems, both simple and complex, will be made up of sensors and knowledge bases with their own ability to communicate and evolve. Signal evaluation based on small samples and pre-processed events available in the machine "cortex" is particularly relevant for processing based on network machines.

In this talk, we will discuss:

• Design of a sample network of sensors and servers: the integration of many simple components that are off-the-shelf, home-grown, and inexpensive. (For example, off-the-shelf robots.)
• Statistically-based algorithms for model estimation.
• Sensor data collection, representation, and visualization.
• Data and system knowledge representation and communication.
• Mobile sensor unit assembly and sensor network configuration, using inexpensive routers, IPcams, light sensors, the Roomba robots, and Linux single-boards.
• Image processing for mobile sensor navigation.

The material to be presented is useful in a variety of domains: research, software implementation, hardware implementation, experimentation with sensor networks, Internet-based toy building, and Internet-based algorithm building.

Jan Smid holds degrees in mathematics, physics, and technical cybernetics from Charles University and Technical University Prague. Jan has consulted for the U.S. government and private industry since 1985. His current research interests include distributed computing (classical and quantum), estimation theory and applications, communication interfaces for sensor networks, knowledge representation, and software and hardware implementations of the networks. All of these components combine for friendly, efficient sensor and database systems of the present and future.

Directions to the Loyola University Maryland Columbia Graduate Center

For more information, please send e-mail to inquiry@cs.loyola.edu.