Systems of Systems Overview

Software engineering emerged in the 1960s to address large-scale, complex software systems' challenges, such as high costs, delays, and unreliability.
The core issue is the increasing complexity and size of systems, which current methods struggle to manage effectively. Modern systems span hundreds to thousands of times the size of older systems, with predictions of systems containing billions of lines of code becoming a reality.
One key factor in handling this complexity is software reuse, which allows developers to integrate existing systems to create large-scale applications.

"System of systems" (SoS) is a collection of two or more independently managed components, distinguishing it from a single system.
This sociotechnical perspective highlights that different parts of an SoS are governed by various management policies, potentially increasing overall complexity.
SoS are relevant in various sectors, but much research in SoS emerged from the defense sector, where independent military systems are integrated for coordinated operations.
Most software SoS involve a limited number of systems now, but their complexity is expected to increase as more systems are integrated to leverage combined capabilities.

Cloud Management System: Manages local and public cloud resources.
Online Banking System: Processes loan requests and connects to credit agencies.
Emergency Information System: Integrates data from emergency services.
Digital Learning Environment (iLearn): Consolidates various software and content.

Operational Independence: Each element in the SoS can function as a standalone system, evolving independently.
Managerial Independence: Different organizations manage elements, leading to varied rules and policies.
Evolutionary Development: SoS evolve over time, not developed as a single project.
Emergence: Unique characteristics of the SoS emerge after its creation, highlighting emergent properties.
Geographical Distribution: Elements often spread across multiple locations, complicating communication and security.
Data Intensive: Software SoS typically manage vast amounts of data, often exceeding the size of constituent systems' code.
Heterogeneity: Various systems within a software SoS are likely developed using different programming languages and methods, reflecting software technology's rapid evolution.

Complexity in software systems, particularly SoS, begins with outlining the parts (elements) and their relationships.
Relationships may include objects, constants, variables, and methods. System complexity is determined by the number and types of these relationships.
Simple systems have fewer relationships compared to complex systems, even with the same number of elements.
Relationships can be categorized as static (analyzed through models like UML) or dynamic (depending on runtime conditions).

System (a) is a relatively simple system with few relationships between elements.
In contrast, System (b), with the same number of elements, is more complex due to many more element-to-element relationships.

Complexity extends beyond the software systems themselves, encompassing development and maintenance processes.
System size increase leads to challenges including difficulty in understanding, coordination issues, and increased documentation requirements.
Increased complexity makes large systems vulnerable to late delivery, cost overruns, and difficulty in understanding and adapting to changes.

Technical Complexity: Arises from relationships among system components, focusing on how they interact and relate to each other.
Managerial Complexity: Related to interactions between the system and its managers, including what managers can change within the system and the relationships among managers overseeing different parts.
Governance: Stems from interplay of laws, regulations, and policies affecting the system, along with decision-making within organizations.

Managerial complexity pertains to operational aspects (what can be done with the system), focusing on practical challenges that managers face.
Governance complexity refers to higher-level decision-making, influenced by national and international laws and regulations that shape policies and impact the system.

Operational Independence, Managerial Independence, and Evolutionary Development contribute differently to varying types of complexity within SoS.
Geographical Distribution and the Data-Intensive nature of systems impact complexity.
Heterogeneity impacts the complexity within a system of systems.

Directed Systems: Single organization owns and develops the system, with a central governing body that sets priorities.
Collaborative Systems: Governed by multiple organizations with shared governance, typically using voluntary governance bodies.
Virtual Systems: Lack central governance, with potential disagreements among participants concerning purpose; systems can enter and exit the SoS.

Organizational Systems: Governed and managed within a single organization.
Federated Systems: Governed by a voluntary body representing all systems' owners.
System of System Coalitions: Lack formal mechanisms, involving informal collaboration, where governance is absent at the organizational level, but informal collaboration exists at the management level.

Reductionism analyzes complex systems by breaking them down into simpler components.
Limitations include emergent properties (behaviors that emerge and cannot be understood by examining components isolated), interdependencies (complex non-linear interactions).
Challenges stemming from oversimplification and dynamic behavior in complex systems, making predictions difficult.

Systems of systems engineering focuses on integrating existing systems to create new functionalities rather than designing from scratch.
Key issues in SoS engineering include limited control, incompatible assumptions, varying evolution strategies, and insufficient support.

Conceptual Design: Creating a high-level vision defines requirements, and identifying constraints informed by existing systems.
System Selection: Choosing relevant existing systems and considering functionality, cost, and governance.
Architectural Design: Developing an overall architecture, ensuring effective interactions.
Interface Development: Creating compatible interfaces enabling system interoperability.
Integration and Deployment: Making systems work together and deploying within organizations.

Direct Communication: Systems communicate automatically with minimal user intervention, ideally through service-based interfaces.
Reconciling Differences: Addresses variability in system interfaces, potentially using service-based interfaces that reflect existing functionalities.
Principal System Coordination: One system acts as a hub directing service calls, managing system interactions.
Interface Development Challenges: Unified user interfaces are often complex and resource-intensive, impacting cost and time.

Integration and deployment are typically separate but should be viewed as unified in SoS to leverage existing systems first, enhancing overall functionality.
Staged deployment can be effective, as demonstrated by the iLearn example.

Testing SoS is challenging due to a lack of detailed requirements, changing constituent systems, and difficulties in resolving potential issues.
Addressing these challenges through agile methods, such as stakeholder engagement and automated testing, is beneficial.