Jad El-khoury

Tool Integration beyond Wasserman

The typical development environment today consists of many specialized development tools, which are partially integrated, forming a complex tool landscape with partial integration. Traditional appr ...

Towards a toolset for architectural design of distributed real-time control systems

We describe a novel tool developed to support architectural design of distributed real-time control systems. The approach enables the co-simulation of functionality (from discrete-time control to logic) together with the controlled continuous-time processes and the behaviour of the computer system. In particular modelling and simulation of distributed computer control systems is supported allowing analysis of timing and control system robustness. An emphasised feature of the tool is its multidisciplinary and integrated approach that combines the views of control and computer engineering into one view at an appropriate level of abstraction. The use of the models and the tool is illustrated through examples and the usefulness of the approach for architectural design is discussed together with avenues for further work. The current models and the tool are designed with the modelling and analysis of fault-tolerant systems in mind. Improved support in this direction is the subject of ongoing work.

A tool integration platform for multi-disciplinary development

In multi-disciplinary development, where various domain specific tools are used by developers to specify and analyze a system, efficient system development requires that the models produced by these tools are well integrated into a whole, reducing any risks of inconsistencies and conflicts in the design information specified. In this paper, we present architecture for a model and tool integration platform that borrows its major components from well known and accepted standards from both computer and mechanical engineering. The architecture supports model integration, where models defined in different tools for different aspects of the same system are related such that they may share and exchange data. The integration platform also enables model management functionalities on a fine-grained level, suggesting a combination of the functionalities found in traditional data management systems such as product data management (PDM) and software configuration management (SCM).

The AIDA toolset for design and implementation analysis of distributed real-time control systems

This article introduces a toolset that integrates the design and performance analysis of control systems with embedded real-time system design. The toolset enables specification and analysis of rea ...

On the modeling and generation of service-oriented tool chains

Abstract Tool chains have grown from ad-hoc solutions to complex software systems, which often have a service-oriented architecture. With service-oriented tool integration, development tools are made available as services, which can be orchestrated to form tool chains. Due to the increasing sophistication and size of tool chains, there is a need for a systematic development approach for service-oriented tool chains. We propose a domain-specific modeling language (DSML) that allows us to describe the tool chain on an appropriate level of abstraction. We present how this language supports three activities when developing service-oriented tool chains: communication, design and realization. A generative approach supports the realization of the tool chain using the service component architecture. We present experiences from an industrial case study, which applies the DSML to support the creation of a service-oriented tool chain. We evaluate the approach both qualitatively and quantitatively by comparing it with a traditional development approach.

Model data management: towards a common solution for PDM/SCM systems

Software Configuration Management and Product Data Management systems have been developed independently, but recently the need to integrate them to support multidisciplinary development environments has been recognised. Due to the difference in maturity levels of these disciplines, integration efforts have had limited success in the past. This paper examines how the move towards model-based development in software engineering is bringing the discipline closer to hardware development, permitting a tighter integration of their data management systems. An architecture for a Model Data Management system that supports model-based development is presented. The system aims to generically handle the models produced by the different tools during the development of software-intensive, yet multidisciplinary, products. The proposed architecture builds on existing technologies from the mature discipline of mechanical engineering, while borrowing new ideas from the software domain.

High-Level Specification and Code Generation for Service-Oriented Tool Adapters

The development of complex systems requires tool support for the different phases of the system life cycle. To allow for an efficient development process, the involved tools need to be integrated, e.g. by exchanging tool data or providing trace ability between the data. Despite the availability of tool integration platforms and frameworks, it is labor-intensive and costly to build tool integration solutions. Industrial tool integration initiatives such as OSLC (Open Services for Lifecycle Collaboration) demand complex configurations and the adherence to integration standards. This further complicates building an integration solution. We propose an approach that uses formalized specifications to systematize tool integration and specialized code generators to automate the process of building tool adapters. We evaluate our approach with the implementation of a code generator that creates service-oriented tool adapters conforming to the OSLC industry initiative.

Qualifying software tools, a systems approach

Modern safety standards designed to ensure safety in embedded system products often take a descriptive approach, focusing on describing appropriate requirements on management, processes, methods and environments during development. While the qualification of software tools has been included in several such standards, how to handle the safety implications of tools integrated into tool chains has been largely ignored. This problem is aggravated by an increase both in automation of tool integration and the size of development environments. In this paper we define nine safety goals for tool chains and suggest a qualification method that takes a systems approach on certifying software tools as parts of tool chains. With this method, software tools are developed and pre-qualified under the assumption that certain properties will be supported by the development environment they are to be deployed in. The proposed method is intended to (1) achieve a stronger focus on the relevant parts of tool chains in regard to safety and (2) separate the extra effort these parts imply from the effort already stipulated by safety standards.

Measuring tool chain interoperability in Cyber-Physical Systems

Cyber-Physical Systems are developed through complex engineering projects that include many stakeholders and a variety of tools and processes from different engineering backgrounds. Interoperability in these development tool chains is an important aspect for well-integrated systems. Furthermore, since full tool chain interoperability is neither possible nor necessarily desired, measuring interoperability in development environments is essential for setting the right priorities. This systematic literature review gives an overview of the literature about interoperability assessment methods. A survey was conducted through digital libraries and a total of 42 papers were read. Out of these papers, 24 different interoperability assessment models were identified and analyzed. A striking find of this study is that no proof of industrial adaptation of these models was found. In this paper, we analyze the reasons for this lack of validation in the context of CPS development.

A Modeling Framework for Automotive Embedded Control Systems

This work is part of a larger effort to define a modeling approach suitable for complex embedded control systems for which modeling techniques are in the progress of maturing. Modeling constitutes an indispensable part of engineering, forming an important basis for documentation, communication, and design automation. In order to describe and compare different modeling techniques, a framework has been developed. The current state of the framework is presented together with lessons learned. We will illustrate the framework with an automotive application. The framework combines a generic concept of systems and the modeling concepts that have been extracted from a study of 12 modeling approaches covering different levels of design and disciplines. Future work will refine the framework by studying additional modeling techniques and by considering more domain specific modeling needs in automotive systems.