Complex System Design with Design Languages: Method, Applications and Design Principles

TL;DR

This paper discusses graph-based design languages in UML that automate complex system design, enabling automatic generation, execution, and feedback of engineering models to optimize product development.

Contribution

It introduces a method using graph-based design languages with a design compiler to automate and accelerate complex system design processes across various industries.

Findings

Successful automation in aerospace, automotive, machinery, and consumer products.

Application of various design strategies and mechanisms for process automation.

Use of similarity mechanics and design patterns to manage complexity.

Abstract

Graph-based design languages in UML (Unified Modeling Language) are presented as a method to encode and automate the complete design process and the final optimization of the product or complex system. A design language consists of a vocabulary (digital building blocks) and a set of rules (digital composition knowledge) along with an executable sequence of the rules (digital encoding of the design process). The rule-based mechanism instantiates a central and consistent global product data structure (the so-called design graph). Upon the generation of the abstract central model, the domain-specific engineering models are automatically generated, remotely executed and their results are fed-back into the central design model for subsequent design decisions or optimizations. The design languages are manually modeled and automatically executed in a so-called design compiler. Up to now, a variety of product designs in the areas of aerospace, automotive, machinery and consumer products have been successfully accelerated and automated using graph-based design languages. Different design strategies and mechanisms have been identified and applied in the automation of the design processes. Approaches ranging from the automated and declarative processing of constraints, through fractal nested design patterns, to mathematical dimension-based derivation of the sequence of design actions are used. The existing knowledge for a design determines the global design strategy (top-down vs. bottom-up). Similarity-mechanics in the form of dimensionless invariants are used for evaluation to downsize the solution for an overall complexity reduction. Design patterns, design paradigms (form follows function) and design strategies (divide and conquer) from information science are heavily used to structure, manage and handle complexity.