Thursday, May 2, 2013
Mechanical and Civil Engineering Seminar
A Wireless Cyber-Physical System Framework for Enhancing the Resiliency of Civil Infrastructure Systems
Jerry Lynch, Associate Professor, Department of Civil and Environmental Engineering , University of Michigan
Earthquakes and other natural hazards remain a serious threat to the safe operation of society's critical infrastructure systems. Further complicating the problem, many of these infrastructure systems are simultaneously approaching the end of their intended design lives with alarming levels of deterioration. Fortunately, the confluence of wireless communications and low-power embedded computing has led to the creation of a new generation of wireless sensing technologies that can be deployed at low cost and in high density to enhance the resiliency of critical infrastructure systems. The paradigm-shift associated with wireless structural monitoring goes well beyond serving as a cost-effective replacement for traditional tethered sensors. Rather, it is the embedded computational intelligence of wireless sensors that transforms them into an autonomous building block of future cyber-physical systems that can be used to monitor and control infrastructure. In this presentation, a wireless cyber-physical system framework is described and illustrated. Validation of the proposed framework is conducted using a permanent wireless monitoring system installed on the New Carquinez Bridge in Vallejo, CA. A dense network of wireless sensors have been installed on the suspension bridge and interfaced to an Internet-based cyberenvironment for the storage and processing of bridge response data. A key element of the proposed framework is an agent-based computational architecture that has been designed to decompose complex data interrogation tasks for ad-hoc distribution within the wireless sensor network itself. The framework is shown to provide a Pareto optimal allocation of scarce network resources (e.g., communication bandwidth, battery power) while maximizing the speed and accuracy of the global algorithm. The presentation concludes with extensions of the framework for real-time feedback control of infrastructure.