A system of systems (SoS) is defined as a set of systems or system elements that interact to provide a unique capability that none of the constituent systems can accomplish on their own. This integration occurs among operationally and managerially independent systems, which retain their distinct identities and purposes while contributing to emergent behaviors in the larger ensemble.[1] Key characteristics distinguishing SoS from traditional systems include operational independence (components function autonomously), managerial independence (separate ownership and funding), geographical distribution, evolutionary development over time, and emergent behaviors arising from interactions rather than individual designs.[2]SoS concepts trace back to early discussions in systems thinking, with one of the first references appearing in 1971 describing collections of organizations as SoS.[3] The term gained prominence in engineering contexts during the late 1990s, particularly through foundational work identifying criteria for SoS architecting.[2] Four primary types of SoS are commonly recognized: directed (centrally managed with tight control over components), acknowledged (formal governance with dedicated oversight of independent systems), collaborative (voluntary partnerships pursuing shared goals without central authority), and virtual (loosely coupled systems appearing integrated through ad hoc interactions). These types reflect varying degrees of integration and control, influencing how SoS are engineered and sustained.[1]Systems of systems engineering (SoSE) addresses the unique challenges of designing, integrating, and evolving such assemblages, often in domains like defense where capabilities emerge from combining existing and new systems.[1] Notable applications include military satellite communications (MILSATCOM) for global connectivity, ballistic missile defense systems (BMDS) integrating sensors and interceptors, and air operations centers coordinating mission planning.[1] Core SoSE elements encompass translating capability objectives into technical requirements, assessing performance against user metrics, developing adaptable architectures, and orchestrating upgrades across asynchronous lifecycles, all while managing risks like emergent behaviors and stakeholder coordination.[1] Standards such as ISO/IEC/IEEE 21839 guide SoSE practices, emphasizing interoperability and resilience in complex, large-scale environments.[4]
Fundamentals
Definition and Characteristics
A system of systems (SoS) is defined as a collection of independent, task-oriented systems that are integrated to deliver unique capabilities unattainable by the individual components alone.[1] These constituent systems pool resources and interact to achieve enhanced overall performance, often emerging from the need to address complex, large-scale problems in domains such as defense and infrastructure.Key characteristics distinguish SoS from traditional systems. Operational independence allows each constituent system to function autonomously, fulfilling its own objectives even without participation in the SoS. Managerial independence means the systems are separately acquired, funded, and managed, with no centralized authority overriding their individual governance.[5]Evolutionary development reflects the dynamic nature of SoS, where the collection evolves over time through incremental changes, upgrades, and adaptations to shifting requirements or environments.[1]Emergent behavior arises from unplanned interactions among the components, producing new properties or capabilities that are not explicitly designed into any single system. Finally, geographic distribution involves components dispersed across wide areas, communicating and coordinating remotely rather than through tight physical integration.[5]In contrast to monolithic single systems, which are designed as cohesive wholes with centralized control and predictable behaviors, SoS exhibit complexity through loose integration of heterogeneous, independent elements, leading to greater adaptability but also challenges in prediction and management.[2] For instance, emergent behavior in a traffic network SoS can manifest as adaptive flow patterns where individual vehicles, operating independently, collectively form efficient congestion relief without a central coordinator.[6]
Types of Systems of Systems
Systems of systems (SoS) are categorized into four primary types based on their structure, governance, and interaction patterns, which significantly influence their design, management, and operational effectiveness. These types—directed, acknowledged, collaborative, and virtual—reflect varying degrees of central authority, constituent system autonomy, and integration mechanisms, as established in foundational systems engineering frameworks.[7][8] This classification helps address how emergent behaviors, a common trait across all SoS, manifest differently depending on the type's coordination level.[9]Directed SoS are centrally managed and engineered to achieve specific, predefined purposes, where constituent systems are often subordinated to the overall SoS objectives. Governance in directed SoS is highly centralized, with a single authority directing development, integration, and operations, allowing for top-down control over resources and changes.[10][8] Integration challenges are relatively minimal, as processes can be enforced uniformly, though maintaining the balance between SoS goals and individual system capabilities requires careful oversight. A representative example is the U.S. Department of Defense's Ballistic Missile Defense System (BMDS), which integrates sensors, interceptors, and command systems under unified management to counter missile threats.[11] In terms of lifecycle implications, directed SoS follow a traditional, hierarchical approach with stable requirements, enabling predictable planning but potentially limiting adaptability to evolving needs.[1]Acknowledged SoS feature recognized objectives, a designated manager, and allocated resources, yet constituent systems retain significant independence in ownership, funding, and operations. Governance operates at a moderate level of centralization, relying on cooperative agreements and persuasion rather than strict authority to align independent entities.[9][8] Key integration challenges include negotiating changes across autonomous systems and ensuring interoperability without overriding individual priorities, which can lead to delays in synchronization. An illustrative case is air traffic management systems, where radar, communication, and navigation components from various providers collaborate under federal oversight like the FAA, while maintaining separate development cycles.[9] Lifecycle management for acknowledged SoS adopts matrix-like structures, supporting evolving solutions through ongoing bilateral coordination rather than rigid control.[7]Collaborative SoS involve constituent systems that voluntarily interact to pursue collectively agreed-upon central purposes, without a dominant central authority imposing decisions. Governance is decentralized, with standards enforced through consensus among key stakeholders, such as industry consortia, fostering cooperation but requiring mutual incentives for participation.[10][8] Integration poses challenges in achieving multi-lateral agreements and sustaining voluntary alignment, particularly when incentives diverge, potentially resulting in fragmented responses during operations. Disaster response networks exemplify this type, where emergency services, NGOs, and local agencies coordinate ad hoc during events like hurricanes, sharing data via common protocols without hierarchical command.[12] The lifecycle of collaborative SoS is dynamic and adaptive, emphasizing customer-focused governance that evolves with situational needs and negotiated changes.[1]Virtual SoS lack a central managementauthority or predefined purpose, with interactions emerging organically among loosely coupled, independent systems driven by market or environmental forces. Governance is absent in a formal sense, relying instead on invisible mechanisms like shared standards or economic pressures to enable coordination among diverse stakeholders.[9][8]Integration challenges are substantial due to high diversity and uncoordinated evolution, making interoperability dependent on voluntary adoption of protocols and increasing risks of misalignment. Global financial markets represent a virtual SoS, where trading platforms, banks, and regulatory bodies interconnect through standardized data exchanges without overarching control, allowing emergent behaviors like market volatility.[13] Lifecycle implications for virtual SoS are emergent and volatile, prioritizing flexible standards for interoperability over top-down planning, which supports scalability but complicates proactive management.[7]
History
Origins and Early Concepts
The concept of a system of systems (SoS) traces its origins to mid-20th century systems science, building on the foundational work of Ludwig von Bertalanffy, who developed general systems theory (GST) in the 1950s as a transdisciplinary framework for understanding complex, interconnected entities beyond isolated components.[14] Bertalanffy's GST, formalized in his 1968 book General System Theory: Foundations, Development, Applications, emphasized open systems with emergent properties arising from interactions among subsystems, influencing early ideas of multi-system integrations in aerospace engineering, such as those explored in the design of large-scale projects like missile and aircraft development during the 1950s.[15] This theoretical groundwork shifted focus from reductionist approaches to holistic views, laying the conceptual basis for later SoS formulations by highlighting the need to manage independence and interdependence in aggregated systems.[16]One of the earliest explicit references to a "system of systems" appeared in Russell L. Ackoff's 1971 paper "Towards a System of Systems Concepts," which described collections of organizations interacting as such systems.[3]In the military domain, initial SoS concepts emerged within the U.S. Department of Defense (DoD) during the Cold War, particularly through the evolution of command, control, communications, and intelligence (C4I) systems in the 1970s and 1980s.[17] These efforts addressed the challenges of integrating disparate, operationally independent platforms—such as radars, communication networks, and surveillance assets—into cohesive networks for strategic operations, driven by the need for enhanced interoperability amid escalating geopolitical tensions.[18] Early C4I initiatives, like those formalized under DoD policies for joint operations, exemplified proto-SoS structures where constituent systems retained managerial autonomy while collaborating toward shared objectives, marking a departure from monolithic designs.[19]Pre-1990s conceptual groundwork further advanced through key contributions in systems engineering literature, including Andrew P. Sage's 1977 work Methodology for Large Scale Systems, which explored hierarchical integrations of complex, distributed entities akin to modern SoS.[20] Similarly, early NASA explorations in space systems during the 1960s and 1970s, such as the multi-component architectures in the Apollo program and advanced concepts outlined in 1976 reports on orbital support needs, applied systems thinking to integrate independent modules like propulsion, guidance, and life support for ambitious missions.[21] These efforts highlighted the practical challenges of autonomy in aggregated systems, informing theoretical distinctions.By the 1990s, key publications began formalizing SoS as distinct from traditional systems engineering, with Mark W. Maier's 1998 paper "Architecting Principles for Systems-of-Systems" providing a seminal taxonomy that differentiated SoS based on operational and managerial independence of components, contrasting them with tightly coupled complex systems.[2]DoD reports from the era, such as those addressing C4I interoperability, reinforced this by emphasizing emergent behaviors in networked defenses over conventional hierarchical engineering, setting the stage for recognition of SoS as a unique paradigm.[17]
Key Developments and Milestones
In the early 2000s, the U.S. Department of Defense (DoD) formalized the recognition of systems of systems (SoS) within its acquisition processes through updates to key directives. The 2003 revision of DoD Directive 5000.1, "The Defense Acquisition System," emphasized flexible approaches to acquiring complex capabilities, including SoS, to address evolving operational needs. Complementing this, NASA's Exploration Systems Mission Directorate (ESMD), established in 2005, developed an SoS framework to integrate human and robotic exploration systems for lunar and Mars missions, focusing on architecture decomposition and interface management.[22]International progress accelerated in the late 2000s with European Union initiatives under the Seventh Framework Programme (FP7). The T-AREA-SoS project, launched in 2012 as part of FP7, advanced trans-Atlantic collaboration on SoS research agendas, particularly for transportation systems, by analyzing economic and societal impacts.[23] Concurrently, the International Council on Systems Engineering (INCOSE) established its Systems of Systems Working Group in the early 2010s to promote SoS engineering applications across domains, building on prior discussions from the late 2000s.[24]The 2010s saw expansions in policy and knowledge formalization. In 2011, the U.S. Government Accountability Office (GAO) issued reports highlighting SoS challenges in DoD acquisitions, such as integration risks and cost overruns in major weapon systems.[25] By 2015, the Systems Engineering Body of Knowledge (SEBoK) version 1.4 incorporated SoS as a dedicated knowledge area, providing structured guidance on SoS characteristics, engineering, and complexity.[26]Recent milestones through 2025 reflect growing emphasis on SoS in environmental and aviation domains via European funding. The EU's Horizon 2020-funded Arctic PASSION project (2021–2025) co-creates an integrated pan-Arctic observing system of systems to address climate challenges, involving Indigenous communities and enhancing data access.[27] Similarly, the Horizon Europe-funded COLOSSUS project (2023–2026) develops a system-of-systems design methodology for aviation, optimizing aircraft, operations, and business models holistically.[28] In 2025, INCOSE's updates, including INSIGHT publications and symposium proceedings, addressed SoS complexity through analyses of resilience and natural language processing for engineering practices.[29]
Timeline of Key Events (2000–2025)
Engineering and Methodology
System-of-Systems Engineering Approach
System-of-systems (SoS) engineering adopts a distinct approach that prioritizes the coordination of multiple independent constituent systems to achieve emergent capabilities, rather than designing a monolithic entity from scratch. Core principles emphasize interoperability to enable seamless data, control, and functionality exchange across systems through standardized interfaces, common data semantics, and message formats. Integration of legacy systems is central, leveraging existing assets at varying maturity levels via techniques such as gateways, middleware, or "glue-ware" to bridge gaps without necessitating wholesale replacements, thereby addressing data mismatches and limitations inherent in older infrastructure. Unlike traditional engineering, which seeks full control over system behavior, SoS engineering focuses on managing emergence—unanticipated behaviors arising from system interactions—through proactive assessment via modeling, simulation, and ongoing monitoring to mitigate risks of harmful outcomes while harnessing potential benefits.[1]The lifecycle of SoS engineering diverges significantly from traditional systems engineering (SE), which typically follows a linear, top-down model with synchronized phases and centralized authority. In contrast, SoS lifecycles are evolutionary and incremental, accommodating asynchronous development cycles among constituent systems managed by separate entities, which demands continuous integration, testing, and adaptation rather than fixed milestones. This federated approach relies on collaboration and loose coupling, where SoS integrators lack direct control and instead coordinate through negotiation, memoranda of agreement (MOAs), or service level agreements (SLAs) to align independent system owners toward shared objectives. Such structures account for the managerial independence of constituents, requiring SoS engineers to facilitate stakeholder negotiations across broader trade spaces, dynamic relationships, and uncertainties, unlike the hierarchical governance of single-system SE.[1]Key processes in SoS engineering include requirements elicitation that spans multiple constituents, translating high-level capability objectives from diverse stakeholders into traceable SoS-level and system-specific needs, often managed iteratively to accommodate evolving inputs. Risk management addresses uncertainties at both SoS and constituent levels, evaluating potential impacts on cost, schedule, performance, and emergent behaviors through ongoing identification, mitigation, and monitoring, with particular attention to technology maturity and integration challenges. Governance structures provide oversight without top-down mandates, employing mechanisms like systems engineering councils, risk boards, configuration control boards, and technical reviews to foster cross-system coordination, decision-making, and policy enforcement among independent parties.[1][30]Prominent frameworks, such as the U.S. Department of Defense (DoD) Systems Engineering Guide for Systems of Systems (2008) and the Systems Engineering Guidebook (2022), underscore modularity and adaptability by advocating open systems architectures— including the Modular Open Systems Approach (MOSA)—persistent overlays for capability evolution, and designs that minimize interdependencies to support extensibility and reconfiguration in response to changing environments. These frameworks highlight the need for SoS engineering to overlay architectures on independently evolving systems, ensuring flexibility while managing interfaces and behaviors that emerge from their interactions.[1][31]
Methodologies and Tools
Model-Based Systems Engineering (MBSE) has been adapted for System of Systems (SoS) to address the complexity of integrating independent constituent systems, emphasizing model-centric approaches that capture requirements, architecture, and interfaces across multiple levels from SoS to components.[32] This adaptation leverages SysML as a foundational language, with extensions such as parametric diagrams and profiles to model interoperability, enabling the specification of shared interfaces and emergent behaviors without proprietary integrations.[33] For instance, SysML v2, released by the Object Management Group (OMG) in July 2025, introduces a standardized application programming interface (API) that facilitates seamless data exchange between SoS models and external tools, supporting virtual integration and reducing development risks in acknowledged SoS where constituent systems retain operational independence.[34][35]Simulation tools play a critical role in SoS engineering by replicating dynamic interactions and emergent behaviors that arise from constituent system collaborations. Agent-based modeling (ABM) is particularly suited for this, as it represents each constituent system as an autonomous agent whose local rules and interactions produce global SoS-level outcomes, such as unexpected resilience or failures.[36] Tools like AnyLogic support hybrid ABM with discrete-event and system dynamics paradigms, allowing engineers to simulate SoS scenarios involving heterogeneous components, such as transportation networks where agent decisions lead to traffic flowemergence.[37] Similarly, MATLAB/Simulink enables ABM through stateflow and SimEvents blocks, facilitating the analysis of time-dependent emergent phenomena in SoS like defense simulations, where agent mobility and decision-making reveal system vulnerabilities.[38]Interoperability standards are essential for enabling distributed SoS simulations and virtual collaborations among loosely coupled systems. The High Level Architecture (HLA), standardized by IEEE 1516-2025, provides a framework for federated simulations where SoS components operate as independent federates, exchanging data via a runtime infrastructure to achieve seamless integration without central control.[39] HLA's object model template and federation object model support interoperability in acknowledged SoS, as demonstrated in defense applications where disparate simulators for air, ground, and naval systems interact in real-time.[40] Complementing this, API frameworks derived from SysML extensions allow for virtual SoS prototyping, where models are queried and updated dynamically to test interface compatibility across distributed environments.[34]Analysis methods in SoS engineering employ network theory to quantify connectivity and resilience, modeling SoS as graphs where nodes represent constituent systems and edges denote interactions. Graph-based metrics, such as betweenness centrality and clustering coefficients, assess resilience by measuring how disruptions propagate through the network, revealing vulnerabilities in interconnected infrastructures like power grids integrated with communication systems.[41] For resource allocation in acknowledged SoS, where central authority is limited, optimization algorithms like deep reinforcement learning frameworks dynamically assign assets—such as bandwidth or personnel—by learning from simulated interactions to maximize overall performance while respecting constituent autonomy.[42] These methods prioritize multi-objective functions, balancing trade-offs in scenarios like collaborative sensor networks.INCOSE endorses computer-aided software engineering (CASE) tools that support SoS visualization and management, with Cameo Systems Modeler standing out for its SysML-based environment that enables hierarchical views of SoS architectures, traceability matrices, and simulation plugins for emergent behavior analysis.[43] This tool facilitates the creation of parametric models for trade studies in SoS, such as optimizing interoperability in multi-domain operations, and integrates with external simulators via OSLC standards for collaborative engineering.[44] Other INCOSE-recommended tools, like those in the Systems Engineering Tools Database, extend this capability by mapping SoS processes to visualization dashboards, aiding in the identification of integration points across distributed teams.[43]
Applications
Defense and Military
In defense and military contexts, systems of systems (SoS) enable the integration of diverse, independently operated components—such as sensors, platforms, and command structures—into cohesive architectures that enhance strategic capabilities across domains. These directed SoS, where constituent systems are managed under a central authority, are particularly suited to military operations requiring unified control and rapid response.[45]A prominent example is the U.S. Department of Defense's (DoD) Joint All-Domain Command and Control (JADC2), which connects sensors and assets from air, land, sea, space, and cyber domains into a unified network to provide decision advantage at the speed of relevance. JADC2 integrates data fabrics, artificial intelligence, and automation to fuse information from disparate sources, enabling commanders to sense, analyze, and act across contested environments. This directed SoS supports globally integrated operations, including with mission partners, by modernizing command and control through resilient networks and enhanced information sharing.[45][46]Intelligence, Surveillance, and Reconnaissance (ISR) networks exemplify SoS applications by linking airborne, ground-based, and space-based platforms to collect and disseminate real-time data for operational awareness. In DoD frameworks, ISR operates as an interconnected architecture that supports joint force planning and execution, filling critical gaps in battlespace intelligence through sensor fusion and dynamic tasking. These networks contribute to broader C4ISR (Command, Control, Communications, Computers, Intelligence, Surveillance, and Reconnaissance) integrations, which optimize data processing from multiple sources to improve situational awareness and targeting precision.[47][48]C4ISR integrations further amplify SoS benefits in defense by fusing vast datasets from global sensors into actionable intelligence, acting as the military's "nervous system" for enhanced operational effectiveness. For instance, these systems enable rapid data sharing across services, reducing decision timelines and increasing resilience against disruptions. In ballistic missile defense, the Ground-based Midcourse Defense (GMD) system integrates ground-based interceptors, space-based infrared sensors, early warning radars, and command nodes to detect and destroy intercontinental ballistic missiles during midcourse flight. GMD's SoS architecture, managed by the U.S. Army's Missile Defense Brigade, provides homeland protection through precise, non-explosive kinetic kills at speeds exceeding Mach 30.[49][50][51]Historical projects in the 2000s, such as Net-Centric Warfare (NCW) initiatives, laid foundational SoS principles by networking sensors, decision-makers, and effectors via the Global Information Grid to achieve information superiority. NCW emphasized shared battlespace awareness and self-synchronization, with implementations like the U.S. Navy's Cooperative Engagement Capability demonstrating improved track accuracy and extended engagement ranges. These efforts promoted enterprise-wide engineering, including System-of-Systems Authorities for interoperability, yielding higher combat tempo and reduced mission risks.[52][53]In the 2020s, hypersonic weapon systems represent evolving SoS applications, integrating boost-glide vehicles, launch platforms, and tracking sensors for long-range, high-speed strikes against time-critical targets. DoD programs like the Army's Long-Range Hypersonic Weapon (LRHW, or Dark Eagle) combine maneuverable payloads with supporting architectures to penetrate advanced defenses, emphasizing survivability and precision over distances up to 1,700 miles at speeds above Mach 5. These systems align with broader modernization strategies, incorporating cross-domain data links for integrated operations.[54][55]Simulations of military SoS have validated benefits, such as the NCW-inspired Fleet Battle Experiment Delta, where linked Army-Navy sensors and shooters reduced threats by an order of magnitude in half the expected time, enhancing response efficiency. Similarly, Air Force data-link integrations in F-15C operations more than doubled kill ratios through faster information sharing, demonstrating resilience and operational gains in joint scenarios. These metrics underscore SoS contributions to quicker decision cycles and sustained effectiveness in complex environments.[53]
Civilian and Infrastructure
In civilian and infrastructure domains, systems of systems (SoS) integrate autonomous, interoperable components to address multifaceted societal needs, such as sustainable urban development, crisis response, and resource management. These applications typically embody collaborative SoS, where independent entities like sensors, networks, and organizations align voluntarily to produce emergent capabilities beyond individual system potentials. By leveraging interconnected technologies, civilian SoS enhance public welfare, economic productivity, and environmental stewardship while adapting to dynamic challenges like population growth and climate variability.Transportation sectors exemplify SoS through large-scale coordination of mobility systems. The SESAR program in Europe functions as an acknowledged SoS for air traffic management, integrating communication, navigation, and surveillance solutions into a holistic framework to modernize performance, reduce delays, and lower emissions across European airspace. Smart city traffic networks further demonstrate SoS principles by interconnecting real-time data from sensors, adaptive signal controls, and predictive analytics to optimize urban flow, as seen in governance frameworks for transportation systems that treat cities as interconnected constituent systems. These approaches enable dynamic adjustments, such as rerouting vehicles based on live congestion data, improving safety and efficiency in densely populated areas.Healthcare and energy infrastructures rely on SoS for resilient, integrated operations amid disruptions. During pandemics like COVID-19, integrated hospital networks operate as SoS, enabling coordinated resource sharing and patient triage across facilities, as evidenced in analyses of public health responses that highlight interdependencies in care delivery. In energy, the U.S. Department of Energy's Grid Modernization Initiative conceptualizes the power grid as a system of systems, where complex components like renewable sources and transmission lines interconnect to bolster reliability and incorporate variable energy inputs. This structure supports emergent behaviors, such as automated balancing of supply and demand, to maintain grid stability during peak loads or outages.Additional sectors underscore SoS for global resilience and observation. Supply chain logistics during the COVID-19 pandemic were framed as SoS, with logistics service providers navigating uncertainties through interconnected networks that revealed cause-effect relationships between disruptions, such as delivery delays and financial risks. Environmental monitoring benefits from the Arctic PASSION project (2021-2025), which co-creates a pan-Arctic observing system of systems to integrate data across land, ocean, atmosphere, and cryosphere domains, providing services like permafrost alerts and pollution forecasts to support Indigenous communities and policymakers.These civilian SoS deliver scalability advantages through emergent efficiencies, including cost reductions in infrastructure via optimized interoperability; for instance, SoS modeling in transportation and energy has enabled operational savings by enhancing resource allocation and minimizing redundancies.
Challenges and Research
Major Challenges
One of the primary challenges in developing and maintaining systems of systems (SoS) is achieving interoperability among constituent systems, particularly when integrating legacy systems with incompatible architectures and protocols. Legacy systems often lack standardized interfaces, leading to significant integration failures and projectdelays; for instance, in the U.S. Department of Defense (DoD), interoperability gaps have contributed to postponed military capabilities and increased risks during operations.[56][57] Standards gaps exacerbate these issues, as varying protocols across independently developed systems hinder seamless data exchange and functionality, requiring extensive retrofitting or middleware solutions that further complicate deployment.[57]Emergent behavior in SoS poses substantial risks due to unpredictable interactions among autonomous components, which can result in system-wide failures not evident in individual systems. These interactions may lead to cascading effects, such as outages in interconnected energy grids where a localized fault propagates rapidly across regional networks, as observed in analyses of major power disruptions.[58] The complexity of modeling and predicting such behaviors is heightened in dynamic environments, where constituent systems evolve independently, making it difficult to anticipate and mitigate adverse outcomes without comprehensive simulation frameworks.)Governance and management of SoS are hindered by the need to coordinate multiple independent owners, each with distinct objectives, funding, and operational priorities, which often results in fragmented decision-making. In virtual SoS, where systems collaborate loosely without central authority, scalability becomes a key issue, as expanding the network amplifies coordination challenges and risks misalignment.[59] Effective governance requires mechanisms for aligning incentives and resolving conflicts, yet the distributed nature of ownership frequently leads to accountability gaps and inefficient resource allocation.)Resource constraints represent a persistent barrier in SoS development, characterized by high costs and extended timelines due to the intricate integration and testing required across disparate components. Acknowledged SoS projects, in particular, commonly experience schedule overruns of several years and cost growth exceeding initial estimates, as evidenced in DoD weapon system portfolios where integration complexities drive up expenditures.[60] These overruns stem from the iterative nature of aligning evolving systems, often necessitating additional funding and personnel without guaranteed returns on investment.[61]Socio-technical factors, including privacy and cybersecurity in distributed SoS, introduce vulnerabilities arising from the interconnected yet decentralized architecture, where data sharing across boundaries heightens exposure to breaches. In such systems, ensuring privacycompliance amid varying regulatory frameworks across constituent owners is challenging, potentially leading to data leaks or unauthorized access that compromise overall integrity.[62] Cybersecurity threats are amplified by the expanded attack surface, as a compromise in one subsystem can propagate through the network, demanding robust, yet often resource-intensive, protective measures like encryption and access controls.[63] The severity of these challenges can vary depending on the type of SoS, such as directed or virtual configurations.)
Current Research Directions
Current research in system of systems (SoS) engineering is driven by persistent challenges such as managing emergent behaviors and ensuring interoperability in dynamic environments.[64]A major focus involves addressing complexity and emergence through AI-driven prediction of system behaviors. Researchers are developing analytical techniques and standard measures to track complexity metrics, predict emergent properties, and mitigate unintended interactions in SoS.[64] For instance, studies emphasize agent-based simulations and formal methods to explore state spaces and correlate parameters for maintaining system health.[64] The Systems Engineering Research Center (SERC), funded by the NSF, advances AI and machine learning applications for data-driven analysis of nonlinear behaviors in SoS and enterprise-scale systems, enhancing prediction capabilities for resilience.[65]Advanced modeling techniques are integrating digital twins and AI to enable comprehensive SoS simulations. The EU-funded COLOSSUS project (2023-2026) exemplifies this by developing a SoSdesign methodology that optimizes aircraft, operations, and business models through a four-level hierarchical approach, incorporating AI-enhanced digital twins for use cases like sustainable intermodal mobility and aerial wildfire fighting.[28] This framework builds on the AGILE design platform to simulate interdependencies and improve aviation system resilience against environmental disruptions.[28]Sustainability-oriented research emphasizes resilient SoS for climateadaptation, particularly in vulnerable regions. The Arctic PASSION project establishes a pan-Arctic Observing System of Systems (pan-AOSS) to monitor environmental changes, co-create integrated observations, and support community adaptation to climate impacts through high-quality Earth observation data. This initiative unifies existing networks to address gaps in long-term monitoring, fostering adaptive responses in Arctic ecosystems.Interdisciplinary approaches are gaining traction by combining systems engineering (SE) with data science to tackle SoS challenges. The 2025 updates to the Systems Engineering Body of Knowledge (SEBoK) highlight mission engineering as a key practice, integrating SE processes with data analytics for SoS capability analysis, architecture modeling using SysML, and iterative simulations to identify gaps and inform investments.[66] These updates, aligned with the US DoD Mission Engineering Guidebook 2.0 (2023), promote quantifiable assessments of mission outcomes in complex SoS environments.[66]As of November 2025, the INCOSE Systems Engineering Vision 2035 (2022) further guides research by emphasizing AIintegration for resilient and sustainable SoS, with recent discussions at the INCOSE International Symposium 2025 reinforcing themes of adaptive systems in response to global challenges like climate change and technological disruption.[67][68]Funding and collaborations underpin these efforts, with significant support from agencies like the NSF and DARPA. The NSF sustains SERC's research on AI-enabled SoS modeling and complexitymanagement through multi-year grants.[65]DARPA's fiscal year 2025 budget allocates resources for SoS warfare architectures, including resilient maritime systems integration.[69] Joint initiatives, such as the IEEE International Conference on System of Systems Engineering (SoSE 2025), co-sponsored by INCOSE, facilitate global collaboration on sustainable SoS development, featuring themes like resilient and adaptive systems.[70]
Education and Professional Practice
Educational Programs
Purdue University offers a Master of Science in Systems Engineering program that incorporates specialized coursework on system of systems (SoS) modeling and analysis, enabling students to address complex, interconnected systems through tools like multidisciplinary design optimization and systems synthesis.[71] The curriculum emphasizes practical application in engineering large-scale systems, preparing graduates for roles in designing and managing SoS environments.[72]The Naval Postgraduate School provides a robust SoS-focused curriculum through its Systems Engineering Analysis Program (Curriculum 308), which trains U.S. Navy officers in building and operating large combat SoS, with a strong emphasis on defense acquisition and lifecycle management.[73] Additionally, the school offers a four-course System of Systems Certificate, designed for applying systems engineering principles to defense acquisition and SoS integration.[74] These programs highlight interdisciplinary approaches, including operational analysis and technological superiority in military contexts.[75]Online and certification opportunities include INCOSE's Systems Engineering Professional (SEP) certifications—Associate (ASEP), Certified (CSEP), and Expert (ESEP)—which cover SoS concepts within broader systems engineering knowledge, supported by the INCOSE System of Systems Working Group for specialized guidance.[76] Complementing these, MIT xPRO's Architecture of Complex Systems course explores SoS-like structures in modern engineering, focusing on minimizing rework in interconnected designs, while Coursera's Introduction to Complexity Science addresses features of complex, connected systems relevant to SoS education.[77]Core curriculum elements across these programs feature courses on Model-Based Systems Engineering (MBSE) tailored for SoS, utilizing digital models to support system lifecycle processes from requirements definition to verification.[78] Instruction often incorporates case studies from Department of Defense (DoD) projects, such as MBSE applications in engineering organizations to enhance process management and integration.[79] An interdisciplinary focus integrates policy considerations, ensuring students understand regulatory and ethical dimensions alongside technical skills.Graduates from programs like those at the Naval Postgraduate School have contributed to key defense initiatives, including research theses on Joint All-Domain Command and Control (JADC2) integration, advancing naval strategies for SoS-enabled operations.[80]
Industry Standards and Practices
The International Organization for Standardization (ISO), the International Electrotechnical Commission (IEC), and the Institute of Electrical and Electronics Engineers (IEEE) have developed ISO/IEC/IEEE 21840:2019, which provides specific guidance for applying the processes outlined in ISO/IEC/IEEE 15288:2015 (Systems and software engineering—System life cycle processes) to systems of systems (SoS). This standard addresses the unique challenges of SoS by emphasizing interoperability, emergent behavior management, and lifecycle processes tailored to collaborative, distributed environments where constituent systems retain operational independence.[81]The U.S. Department of Defense (DoD) issued the Systems Engineering Guide for Systems of Systems in 2008 (version 1.0), offering practical guidance for integrating independently managed systems into larger SoS architectures, with a focus on acknowledged SoS that have defined objectives and dedicated management. This guide outlines core engineering elements such as translating capability objectives, assessing constituent system performance, and orchestrating incremental upgrades, while referencing related DoD policies like the 2004 Systems Engineering Policy for risk and interface management. Although no dedicated SoS guide update occurred in 2023, the broader DoDSystems Engineering Guidebook was revised in 2022 to incorporate evolving practices, including SoS considerations in adaptive acquisition frameworks.[1][82]The International Council on Systems Engineering (INCOSE) plays a pivotal role through its Systems of Systems Working Group (SoSWG), established to advance SoSengineering practices across domains by developing guidance, bibliographies, and reports on challenges like pain points in integration. The group fosters real-world adoption by partnering with organizations such as IEEE and the National Defense Industrial Association (NDIA), influencing standards like ISO/IEC/IEEE 15288 and contributing to the Body of Knowledge and Knowledge Areas for Systems Engineering (BKCASE) with diverse practitioner input. Boeing employs a standardized System of Systems Engineering (SoSE) process that leverages model-based approaches to architect complex integrations, as demonstrated in defense programs where requirements are traced through digital models to ensure extensibility and minimize interdependencies. Lockheed Martin utilizes mission integration protocols, including the Integrated Combat System and the STAR.OS framework, to unite platforms, networks, and AI-driven autonomy in SoS environments, enabling scalable, netted combat management with common software infrastructure.[83][84][85][86][87]Best practices for acknowledged SoS emphasize risk-based governance to navigate organizational boundaries and independent constituent systems, as detailed in the DoD guide, which recommends establishing integrated risk management boards, assessing technology maturity, and prioritizing risks in capability objectives and architecture changes through stakeholder collaboration and memoranda of agreement. For evolutionary development, industry adapts agile methods by focusing on incremental upgrades and loosely coupled architectures, allowing asynchronous system enhancements via regular delivery points (e.g., quarterly "bus stops") and ongoing design analysis to accommodate dynamic environments, as promoted in INCOSE's agile systems engineering initiatives and the DoD's orchestration of upgrades. These practices balance traditional systems engineering with flexibility, using tools like critical path analysis and regression testing to verify emergent behaviors without disrupting legacy systems.[1][88][89]In aviation, Airbus implemented model-based systems engineering (MBSE) for the A350 XWB program, integrating electrical, hydraulic, and avionics subsystems through simulation-driven processes to achieve simplified architectures and reduced maintenance, exemplifying SoS principles in fleet-level operations where multiple aircraft systems interoperate dynamically. In the energy sector, the Consortium of Hybrid Resilient Energy Systems (CHRES), supported by the U.S. Department of Energy's National Energy Technology Laboratory, exemplifies collaborative SoS implementation by hybridizing renewables, storage, and grid components into resilient frameworks, as seen in projects on DC microgrids and offshore wind integration that model interdependent energy flows for enhanced reliability.[90][91]As of 2025, trends highlight the adoption of AI standards to enhance SoS interoperability in supply chains, with initiatives like the NIST Plan for Global Engagement on AI Standards promoting consensus-based frameworks for secure data exchange and model signing to build trust across distributed networks. In military and commercial contexts, AI-driven tools are integrating predictive analytics and autonomous decision-making into supply chain SoS, reducing logistics costs by 5-20% through standardized interoperability that enables seamless subsystem coordination, as evidenced by the Defense Logistics Agency's strategic priorities for digitalAI solutions.[92][93][94]