Systems Design Overview

Principles

Definition

Systems design is an academic concept and methodological approach concerned with the structured creation of complex systems. It investigates the principles and processes for defining system architecture, components, interfaces, and data, emphasizing the holistic integration and optimization of elements to meet requirements and constraints throughout the system lifecycle.

Ontological type

Core Methods

Architecture Frameworks

Requirements Analysis

Key developments

Key Figures

Viewpoint-Driven Requirements era

Axel van Lamsweerde's KAOS framework (early 1990s) introduced goal-oriented requirements engineering that links stakeholder goals to concrete requirements, identifies obstacles, and establishes traceability from goals to specifications. Ian Sommerville, in his 1990s software engineering texts, helped codify structured requirements engineering practices and the use of stakeholder viewpoints to align functional and nonfunctional goals. Philippe Kruchten's 4+1 view model (mid-1990s) operationalized viewpoint-driven design by organizing architecture into multiple synchronized views—logical, development, process, and physical—facilitating communication across stakeholders and supporting traceability. Together these strands stabilized core vocabularies and practices for viewpoint-driven requirements, shaping stage-based workflows, traceability techniques, and responsibility partitioning among design actors in education, standards, and tooling.

Architecture-Centric Design era

Mike Bourne [1] engaged with the University of Cambridge [3] and Cranfield University [4] during the Architecture-Centric Design era. His key contribution was the paper Performance measurement system design: developing and testing a process-based approach [5], which proposed a process-based framework to align metrics with organizational workflows and architectural decisions, a crucial capability for architecture-centric design in this era. Andy Neely [2] helped develop the Performance measurement system design: developing and testing a process-based approach [5], sharpening process-based metrics to support strategic architectural decisions and cross-disciplinary collaboration. Neely [2]'s approach, as articulated in the paper Performance measurement system design: developing and testing a process-based approach [5], enabled lifecycle-aware evaluation that supported scalability, interoperability, and cross-disciplinary collaboration in architecture-centric design.

Model-Centric Digital Twins era

Jakub Konečný [1], connected with University of Edinburgh [3] and Comenius University Bratislava [4], exemplifies the model-centric orientation of the era. Jakub Konečný [1] contributes through the paper Towards Federated Learning at Scale: System Design [7], which outlines system design principles for scalable, privacy-preserving federated learning and enables cross-institution model-driven exploration. David Petrou [2], associated with Wuhan University of Technology [5] and University of Electronic Science and Technology of China [6], also typifies this era. His key contribution, reflected in Towards Federated Learning at Scale: System Design [7], emphasizes scalable, privacy-aware federated learning and its role in enabling distributed computation for model-centric digital twins.