DDDAS, A Key Driver for Large-Scale-Big-Data and Large-Scale-Big-Computing

This talk will provide an overview of future directions in Big Data and Big Computing, as driven by the DDDAS paradigm. In DDDAS,the computation and instrumentation aspects of an application system are integrated in a dynamic feed-back control loop. Thus, by its inception DDDAS is a driver for application support environments where the computational platforms span the diverse range of high-end and mid-range computing and including the instrumentation platforms, stationary and mobile networked sensors, and end-user devices. Commensurately, the data involved in DDDAS environments span data associated with complex computer models of application systems to instrumentation-data –either collected from large instruments or from the multitudes of heterogeneous mobile and stationary ubiquitous sensing devices, and including end-user devices. Data from ubiquitous sensing constitute the next wave of Big Data. These collections of ubiquitous and heterogeneous sensing devices not only are sources of large volumes of heterogeneous sets of data, but also the amount of computing that is performed collectively on the multitudes of instrumentation/sensor platforms amounts to significant computational powerwhich should be viewed in tandem with that performed in the high-end platforms. In DDDAS environments, this range of platforms - from the high-end, to the instrumentation and end-user platforms - constitute the dynamically integrated, unified platform referred to here as Large-Scale-Big-Computing (LSBC); the diverse sets of data - from high-end computing data to data from large sets ofheterogeneous sensing are referred to as Large-Scale-Big-Data (LSBD). There are challenges and opportunities in supporting and exploiting Large-Scale-Big-Computing and Large-Scale-Big-Data. DDDAS has been applied and is creating new capabilities in many application areasspanning systems from the nanoscale to the terra-scale and the extraterra-scale, andcovering a multitude of domains such as:physical, chemical, biological systems (e.g.: engineered materials, protein folding,bionetworks); engineered systems (e.g.:structuralhealth monitoring,decision support andenvironment cognizant operation); surveillance, co-operative sensing,autonomic coordination, cognition; energy efficient operations; medical and health systems (e.g.: MRI imaging, seizure control); ecological and environmental systems (e.g.:earthquakes, hurricanes, tornados, wildfires, volcanic eruptions and ash transport,chemical pollution); fault tolerant critical infrastructure systems (e.g.: electric-powergrids, transportation systems); manufacturing processes planning and control; space weather and adverse atmospheric events; cybersecurity; systems software. DDDAS is a driver for LSBC and LSBD environments and but also a methodology to efficiently manage and exploitthese large-scale-heterogeneous resources, aspects which will be addressed in the talk; additional examples of DDDAS-based new capabilities for such applications are provided in other papers in this workshop