Computing for Distributed Decision Support and Information Coordination Hubs in a Pervasively Connec
Abstract
PUBLICALLY RELEASABLEONR PO Contact: Predrag NeskovicWe request support for an expanded computational system (ECS) to bolster our on,going DoD work (specifically with the DTRA) to augment and support their response and planning efforts in the event of large scale d,isasters in the post-COVID world. State-of-the-art computational foundations and system engineering are essential in sustaining DoD,and DTRAs missions to avert, mitigate, and respond to human-initiated and natural global/societal events such as pandemics, large-s,cale urbanization, migration caused by societal unrest and shifting resources and opportunities, and cascading failures across multi,ple infrastructures caused by WMD,cyber-attack, or a natural event such as a hurricane. The COVID-19 pandemic has created an unprece,dented disruption to society, and developing decision support tools in this new normal is the primary focus of our work. Proliferati,on of increasingly capable distributed or mobile computing devices and extensive selective/timely data accessibility, their uses and, consequences during complex disasters and periods of societal challenge, have changed the diverse problems of governance. From situ,ation assessment to large scale planning to detailed coordinated control of facilities and resources, governance approaches have lag,ged the emerging realities -- theCOVID-19 pandemic has made this vividly clear. Relevant integration of distributed data analytical,systems and models across these scales in real-time decision making will require layers of relatively more central compute hubs th,at are not necessarily major HPC, cloud, or data centers. The ECS will enhance the Network Systems Science and Advanced Computing (N,SSAC) division of the Biocomplexity Institute (BI) in crisis management capability in support of DTRA Reachback and other DoD organi,zations in the post-COVID world. The system will also support related functions, such as policy analysis, planning, course-of-action, analysis, incidentmanagement, and training in a variety of domains (e.g., urban evacuation management, epidemiological event manage,ment, bio-monitoring, population risk exposure estimation, logistical planning and management of isolated populations, site evacuati,ons, interdependent infrastructure failures). Through modeling and analytics enabled by the ECS, DoD analysts and scientists will be, able to explore new research questions about societal infrastructures and the individuals interacting with them, allowing policy ma,kers and emergency managers unprecedented opportunities for agile integration of converging information streams. We propose the expa,nsion of the computational system to augment our existing HPC cluster, the University of Virginias Rivanna cluster. This seamless i,ntegration with our current cluster willmaximize the reuse of existing equipment (storage, networking, management) while at the same, time increasing both the throughput that can be achieved with the system, as well as the sizeand complexity of simulation and analy,sis that can be performed. The resources can be dedicated in support of DTRA Reachback to provide for high throughput and real-time,needs, extending the value of resources currently employed.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 01, 2022
- Source ID
- N000142212273
Entities
People
- Christopher Barrett
Organizations
- Office of Naval Research
- United States Navy
- University of Virginia