Semantic Foundations and Formal Methods for Evolutionary System-of-System Architectures
Principal Investigator: Mark A. Austin, University of Maryland, College Park
Co-Investigators: Jennifer Golbeck, University of Maryland, College Park
Years of Award: 2023-2026
Managing Service Agency: Army Research office
Military systems are currently burdened with a multiplicity of factors that prevent their ability to rapidly deploy new technologies and integrate those technologies across programs to a portfolio level. Rigid management structures, services rivalries, national and local politics, aging management processes, and lack of digitalization are factors contributing to inferior levels of performance and agility in present-day systems. Looking ahead, these sources of complexity point to a strong need for new approaches to understanding, modeling, synthesizing and optimizing the operations of large-scale, distributed, institutions and organizations (so-called system-of-systems). With expertise spanning economic and geopolitical modeling of institutions, information and trust modeling, and systems-of-systems modeling, our multidisciplinary team of researchers will explore the benefits of semantic foundations and formal methods for the synthesis and formal analysis of evolutionary system-of-system (SoS) decision models. To achieve and demonstrate the core objectives of the proposed research, the plan of work will pursue a coordinated agenda organized into two themes: (1) Semantic Foundations for Institutional Governance and Organizational Trust, and (2) Semantic Foundations, Compositional Behavior Modeling, and Formal Analysis of Evolutionary System-of-System Architectures. The first theme will contribute to the foundations of social science through: (1) Semantic foundations for institutional governance (i.e, rules and culture, along with their spatial and temporal extent, and history of enforcement), and (2) Semantic foundations for the personal and organizational aspects of trust. The second theme will address the challenges of designing system-of-system architectures that need to dynamically respond to external influences, and in a manner that mediates unnecessary friction (e.g., failure caused by untrustworthy relationships) and prevents undesirable emergent properties and behaviors. The research will be divided into three tasks: (1) Semantic foundations for networks of organizational interfaces, (2) Compositional behavior modeling and formal methods for evolutionary SoS architectures, and (3) Explore strategies for enhanced performance and agility of DoD Decision Support Systems. A fundamentally new aspect of our proposed research is linkages of institutional governance and networked organizational behavior modeling to formal approaches to compositional behavior modeling and evaluation evolutionary system-of-system architectures. The proposed program of research aligns with the 2022 National Defense Strategy priorities focus on readiness, responsiveness, and resiliency of the defense system by “building a resilient Joint Force and defense ecosystem.” If successful, deliverables from this project will take a significant step toward enhanced system capabilities that are interoperable, reconfigurable, and traceable to meet new system threats. Our research will provide the ability to imagine and explore alternative institutional designs including organizational structures (ontologies), their incentives (rules, laws, norms), and the data used for decision making (measures that matter) that can achieve agility and, conversely, do not in the context of uncertainty. Dynamic networks of multi-domain semantic graphs will be mapped to Joint Forces defense operations, thereby providing a framework for data-driven decision making at the speed of relevance.
The project PI is Professor Mark Austin, Department of Civil and Environmental Engineering (CEE) and Institute for Systems Research (ISR), University of Maryland. Dr. Austin is an expert in semantic modeling and reasoning, and formal methods for system validation/verification. Professor Jennifer Golbeck, College of Information Studies, is an expert in modeling of trust relationships. She will serve as project co-PI. Professor John Wallis, Mancur Olson Professor of Economics, will provide expertise in the modeling and development of rules in institutions. For the data and case study application components of the proposed work, the project will draw on the program manager expertise (Defense Advanced Research Projects Agency) and DoD research experience of Dr. Stuart Milner, Research Professor, CEE. The research team will identify new directions for inquiry, disseminate their findings to the broader community, and work with graduate students at the intersection of systems modeling and social science.