What Are Systems Of Systems?

Systems of systems (SoS) are collections of independent, operationally and managerially distinct systems that are integrated to obtain capabilities that no constituent system could achieve alone. Each constituent system retains its own management structure, user base, and operational purpose while simultaneously contributing to the behavior of the larger SoS. This dual nature distinguishes an SoS from a traditional complex system: a conventional system is designed as a unified whole with centralized governance, whereas an SoS is typically assembled from pre-existing systems that were developed independently and continue to evolve independently throughout their service lives.

The concept was formalized by Maier (1998) and elaborated in subsequent INCOSE and IEEE literature. The SEBoK article on System of Systems and Complexity identifies four SoS archetypes defined by the degree of centralized authority: virtual (no central authority), collaborative (voluntary coordination), acknowledged (recognized objectives with limited authority), and directed (centrally managed constituent systems that nonetheless retain some operational independence). The Internet, the US National Airspace System, and combined joint military operations are frequently cited as examples spanning this spectrum.

Characteristics and Engineering Challenges

The key characteristics of an SoS include operational independence of constituents, managerial independence, evolutionary development, emergent behavior, and geographic distribution. Emergent behavior, the appearance of capabilities or failures at the SoS level that cannot be predicted from the behavior of individual constituent systems in isolation, is both the primary motivation for SoS formation and the primary source of engineering risk. SoS engineers work with constrained authority: because constituent systems retain their own governance, the SoS engineer typically influences rather than controls. Requirements at the SoS level must be decomposed into contributions expected from each constituent, but those contributions may change as constituent systems are upgraded, replaced, or retired. Analysis methods from systems engineering, including trade-off studies and effectiveness modeling, must be adapted to account for this evolutionary and distributed governance structure.

Safety and Security

System safety and system security are particularly demanding in the SoS context because threats and hazards can propagate across constituent systems in ways that no single system's safety case anticipates. A failure in one constituent, or a vulnerability exploited in one interface, can cascade into emergent failures across the SoS. Safety analysis for SoS extends classical hazard analysis methods, such as fault tree analysis and FMEA, to cover cross-system interactions and the emergent behaviors they produce. Security analysis must address the expanded attack surface that results when systems from different vendors, procured at different times and under different standards, are connected at runtime. The INCOSE Systems Engineering Lecture materials on system safety and security address how safety and security cases are structured for programs operating at the SoS scale.

Prognostics and Health Management

Prognostics and health management (PHM) in an SoS context extends individual-system condition monitoring to cover the health of inter-system interfaces and emergent SoS capabilities. PHM systems collect sensor data, operational logs, and diagnostic reports from constituent systems and integrate them to assess SoS-level readiness. Predictive models estimate when constituent systems or interfaces will degrade below thresholds that would impair SoS mission performance, enabling maintenance to be scheduled before failures propagate. This requires shared data architectures, agreed health metrics, and governance processes that span multiple system owners, all of which present coordination challenges beyond those of single-system PHM. The IEEE SMC Society's Systems of Systems technical page covers ongoing research into SoS engineering methods including health management.

Applications

Systems of systems engineering is applied across domains where large-scale integration of independent systems is required, including:

• Defense and military operations, combining sensors, platforms, command systems, and logistics networks
• National air traffic management, integrating radar, navigation, and control systems from multiple operators
• Smart grids, where generation, transmission, distribution, and demand management systems are federated
• Disaster response coordination, linking emergency services, communication networks, and logistics systems
• Autonomous transportation networks integrating vehicles, infrastructure, and traffic management systems