EECS Fall 2014 Department Seminar

Richard Kolacinski
Information-Centric Sensor Networks for CPS
White Building, Room 411
11:30 AM - 12:30 PM
October 14, 2014

The increasing availability of inexpensive sensors, communications, and microprocessors with decreasing footprints is providing the ability to collect, share, and process data at heretofore unobtainable rates and levels of resolution is thus engendering new design goals, objectives and paradigms for everyday systems.  Common objectives for these Cyber-Physical Systems (CPS) include imbuing them with resiliency/self-healing, flexibility/adaptation, and enabling improved visibility, operational performance, automation/control, and decision making.  A crucial component to realizing these objectives the development of appropriate theoretical frameworks, tools, and techniques for assimilating the data and eliciting the actionable information required.  To this end, an information-theoretic framework is introduced wherein the constituent elements of the CPS are treated as information processors and the various physical phenomena associated with the CPS elements are viewed as communication signals. 

Noting that information is distinct from data, data that tells you what you already know is not informative hence contains no information, an information-centric perspective provides an explicit mechanism for filtering and compressing data.  Further, the emphasis on information rather than data permits the innate organization of the system to be elicited via the properties of information.  Knowledge of this intrinsic structure provides the basis for associating information sources with the appropriate monitoring, decision making, and control processes.  Next, the deep connection between information and thermodynamics is exploited to construct “summary variables” for characterizing large-scale complex systems and constructing behavioral models that capture the system’s evolution at useful mesoscopic levels of resolution.


Dr. Kolacinski is a Research Associate Professor in the Department of Mechanical and Aerospace Engineering where he is performing research on nonlinear dynamical systems, stochastic systems, information theory, and complexity theory and their application to monitoring, event detection, model identification and estimation, and decision and control systems for energy systems.  Prior to joining CWRU, Dr. Kolacinski served as the Technology Lead for Smart Grid Technology at the C.S. Draper Laboratory, where he was responsible for research and development in the areas of control, filtering, system identification, and probabilistic model inference for complex power systems, and as the Director of Advanced Systems for Orbital Research, Inc. where he performed research on nonlinear and biologically inspired control of unmanned vehicles,  active flow control, and advanced stochastic filtering techniques for navigation and system identification.  Dr. Kolacinski has authored and co-authored over 40 research papers on nonlinear and complex systems, nonlinear and biologically inspired control, filtering, and estimation, and intelligent system design.