ABSTRACT Model Based System Engineering is considered today as the approach that can meet the continually growing complexity of avionics, a challenge that is compounded by constant market pressure (cost, time to market, need for product variants…) For each activity in the product life cycle, tools and technologies supporting an MBSE approach already exist, such as embedded code generation, formal safety analysis, electrical harness design; however, one of the greatest challenges consists in integrating these diverse system development tools into a global framework that ensures the consistency of perspectives and a seamless workflow across processes. Citrus is being developed as an open environment that targets the main activities of systems and software engineering: modeling and validating functions, allocating functional and non-functional requirements to systems, developing the physical architecture, interface design, allocating system requirements to software and hardware items. The Citrus approach relies on key features: efficient asset management, avionics architecture modeling languages that build on the best aspects of SysML and AADL, a rich and user-friendly architecture editing environment that facilitates the use of modeling languages, an efficient extension mechanism that makes it possible to plug specialized external tools (e.g. electrical wiring tool and software application model compilers). Finally, through early testing and analysis, requirements will be validated as they evolve and their partial implementation is verified. This paper also highlights the expected benefits of applying Citrus to a real cockpit application, guided by ARP 4754A and DO-178C.INTRODUCTION At the time of writing, Citrus is in the early stages of development. This paper is mostly focused on presenting the ideas and the related concerns that arise when it comes to developing a new System and Software Engineering environment. Many different approaches (processes, tools and notations) have already been published about the activities involved in engineering complex systems: requirement engineering, architecture description, life-cycle management, test management, etc. It is now a real challenge to select and choose, for each process area, a good candidate among the different approaches so that the global solution is consistent, provides a seamless workflow across processes and is extensible. This paper describes our choices for the first steps of the Citrus roadmap and gives an overview of what will follow and what choices we will be facing. Citrus is designed for two classes of projects: 1. Aircraft functions simulation, where the goal is to configure and simulate a system-of-interest assembled from components provided by multiple suppliers - typically a group of heterogeneous equipment and applications with a simulation of the aircraft and its backbone databus. These projects typically reuse subsystem simulators in various versions from a library of components. Citrus can facilitate the automatic composition of these configurations. 2. Actual avionics equipment design, both Integrated Modular Avionics and federated architectures. In this context, Citrus can assist in validating the stated composition of a configuration. Citrus: Model-Based Avionics Development with Zest!2013-01-2178 Published 09/17/2013 Benoit Viaud ARTAL Technologies Pierre Labrèche Esterline CMC Electronics Copyright © 2013 SAE International doi:10.4271/2013-01-2178Downloaded from SAE International by University of British Columbia, Wednesday, August 01, 2018INITIAL FOCUS The Citrus roadmap was built trying to find a balance that would start with pivotal choices and providing added-value to the end-users as soon as possible. We have decided to start with two major features that are detailed below: • Asset management : as a feature that may be considered as orthogonal to the system development process, it's not an obvious candidate to start with. But because it stru