Systems Architecture Modeling with the Arcadia Method

A Practical Guide to Capella

Pascal Roques

• Publisher: ISTE Press - Elsevier
• Year: 2017
• Print ISBN: 9781785481680
• eBook ISBN: 9780081017920
• Topic: Computer Science
• Subtopic: Hardware
• Index: Computer Science

Systems Architecture Modeling with the Arcadia Method is an illustrative guide for the understanding and implementation of model-based systems and architecture engineering with the Arcadia method, using Capella, a new open-source solution.More than just another systems modeling tool, Capella is a comprehensive and extensible Eclipse application that has been successfully deployed in a wide variety of industrial contexts. Based on a graphical modeling workbench, it provides systems architects with rich methodological guidance using the Arcadia method and modeling language. Intuitive model editing and advanced viewing capabilities improve modeling quality and productivity, and help engineers focus on the design of the system and its architecture.This book is the first to help readers discover the richness of the Capella solution.- Describes the tooled implementation of the Arcadia method- Highlights the toolset widely deployed on operational projects in all Thales domains worldwide (defense, aerospace, transportation, etc.)- Emphasizes the author's pedagogical experience on the methods and the tools gained through conducting more than 80 training sessions for a thousand engineers at Thales University- Examines the emergence of an ecosystem of organizations, including industries that would drive the Capella roadmap according to operational needs, service and technology suppliers who would develop their business around the solution, and academics who would pave the future of the engineering ecosystem

1

Reminders for the Arcadia Method

Abstract

System engineers have been making use of modeling techniques for a long time. Structured analysis and design technique (SADT) and structured analysis for real time (SA/RT) are some of the best known of these, and date back to the 1980s. There are many other approaches based on Petri nets or finite state machines. However, these techniques are also limited by their range and expressivity, as well as by the difficulty in integrating them with other formalisms and with requirements.

Keywords

Arcadia; Computer software configuration item (CSCI); End Product Breakdown Structure (EPBS); Functional Chain; Functional Exchanges; Logical Components; Structured analysis and design technique (SADT); SysML language

1.1 Novelties, strengths and principles

1.1.1 History

System engineers have been making use of modeling techniques for a long time. Structured analysis and design technique (SADT) and structured analysis for real time (SA/RT) are some of the best known of these, and date back to the 1980s. There are many other approaches based on Petri nets or finite state machines. However, these techniques are also limited by their range and expressivity, as well as by the difficulty in integrating them with other formalisms and with requirements.

The rise of UML [ROQ 04] in the world of software and the industrial effort toward developing the tools that accompany it have naturally led to its use being considered in system engineering. However, due to a design process that was strongly influenced by its intended use in object programming, the language was, at least in the early versions, not particularly adapted to modeling complex systems, and was therefore not well suited to system engineering.

An interesting attempt was the publication of a UML variant for system engineering in 2006–2007. This new language, called SysML [CAS 18], was strongly inspired by version 2 of UML, but added the possibility of representing system requirements, non-software elements (mechanical, hydraulic, sensors, etc.), physical equations, continuous flows (matter, energy, etc.) and allocations. Unfortunately, in practice it has been shown that the filiation of the SysML language to UML often leads to difficulty in terms of comprehension and use for system engineers who are not also computer scientists.

This is the reason that led Thales to define the Arcadia method [VOI 16, VOI 17], along with its underlying formalism, for its own needs. It has been applied since 2011 in a growing number of projects across a great variety of domains (avionics, railway systems, defense systems in all fields, air traffic control, command control, area surveillance, complex sensor systems, satellite systems and ground stations, communications systems, etc.), and in many countries (France, Germany, United Kingdom, Italy, Australia, Canada, etc.).

The modeling aspect of the method is supported by a dedicated tool that responds to the constraints involved with full-scale application in an operational context. This tool, called Capella (Melody Advance internally at Thales), is currently freely available for the engineering community as an Open Source application.

1.1.2 Founding principles

Today’s complex systems are limited by a number of requirements or constraints, often concurrently, and sometimes contradictorily: functional requirements (services expected by the users), and nonfunctional requirements (security, operating safety, mass, scalability, cost, etc.). The initial engineering phases of such systems are critical as they condition the aptitude of the architecture used to answer the needs of the clients, as well as the proper distribution of the requirements toward the components, arising from the architecture used. In order to properly handle delays and costs, it is vital to be able to verify the adequacy of the solution with regard to needs from the system design phase, and to minimize the risk of coming across limitations of the solution – thus jeopardizing the architecture – at a more or less advanced stage of development, or even during integration or qualification of the system.

Current practice in systems engineering in the last few years has made use (and rightly so) of a formalization of needs and expectations expressed by the client, in the form of textual requirements, which are then traced (manually) during realization to justify them in relation to the client needs. The limitations of this approach arise mainly from the fact that non-formalized, textual requirements make it harder to verify their coherence and their completeness. Moreover, they are confined to the expression of need and are therefore poorly adapted to describing the solution and to mastering its complexity, or to structuring the engineering. This is one of the reasons that led Thales to the development and deployment of an innovative approach called Arcadia.

Arcadia is a structured engineering method aimed at defining and validating the architecture of complex systems. It favors collaborative work between all stakeholders – of which there are often many – involved in the engineering (or definition) phase of the system. It allows for iterations to be carried out from the definition phase that will help converge the architecture toward congruence with all of the needs identified.

Textual requirements are still present and used as a main contribution toward expressing need at the start of the engineering process. As such, Arcadia takes its place as a major support for engineering and its control, relying on a formalization of the analysis of need, whether operational, functional, or non-functional (functions expected of the system, functional chains, etc.), and on the definition/justification of the architecture based on this functional analysis.

The general principles of Arcadia are the following:

– all of the engineering stakeholders share the same methodology, the same information, the same description of the need and the product in the form of a shared model;

– each specialized type of engineering (for example security, performance, cost and mass) is formalized as a “viewpoint” in relation to the requirements from which the proposed architecture is then verified;

– the rules for the anticipated verification of the architecture are established in order to verify the architecture as soon as possible;

– co-engineering between the different levels of engineering is supported by the joint elaboration of models, and the models...

Table of contents

• Cover
• Title page
• Table of Contents
• Copyright
• Foreword
• Preface
• 1: Reminders for the Arcadia Method
• 2: Capella: A System Modeling Solution
• 3: Complete Example of Modeling with Capella: Operational Analysis
• 4: Complete Example of Modeling with Capella: System Analysis
• 5: Complete Example of Modeling with Capella: Logical Architecture
• 6: Complete Example of Modeling with Capella: Physical Architecture
• 7: Complete Example of Modeling with Capella: EPBS
• Conclusion: Capella’s Key Strengths
• Bibliography
• Index