3rd International ICSE workshop on Software Engineering for Automotive Systems Shanghai, May 23rd, 2006 www.infsec.ethz.ch/events/seas06/

• The 4th International ICSE workshop on Software Enginieering for Automotive Systems will take place in Minneapolis, May 2007
• Workshop summary available
• Slides of the talks available
• Slides of the keynote presentation available
• Program and abstract of invited talk available
• Abstracts available
• Deadline for camera-ready copies extended to March 14th.
• Michael Reinfrank (Siemens VDO) agreed to deliver the keynote
• Workshop Summary 2nd workshop in St. Louis, 2005: pdf: 80K
• Workshop Summary 1st workshop in Edinburgh, 2004: pdf: 80K

Call for Papers [pdf]

Both academia and industry are addressed to submit high-quality papers on approaches relevant to automotive software engineering. Submitted papers might focus on innovative specification languages, case studies, and empirical investigations. The collaboration between university and industry, the industrial applications of research results, and possibly commercial tool support are of equal interest and will be discussed at the workshop. Specific workshop topics are the following.

• Specification Languages
• HW-/SW-Reference-Architectures
• Requirements Engineering
• Model-based SW-Engineering
• Component and Middleware for Embedded Systems
• Code Generation
• Process Metrics
• Quality Assurance, Test Strategies, Verification, Compatibility

The role of software in automotive electronics is in a process of tremendous change. While in the early years of electronic control units for cars, both their function and the related cost of development and manufacturing was dominated by hardware, today the cost of algorithm and software design by far outweigh the cost of hardware. Looking at the material and R&D cost for a microprocessor core compared to the software for control algorithms "in-side", we are facing a 25% HW vs. 75% SW cost ratio for a modern engine control unit. Unfortunately, also the non conformance costs for recalls related to SW issues have grown significantly over the past decades. Therefore, the capability to design SW solutions in a high quality, high performance, and timely to market way has become one of the decisive competitive factors in the automotive supplier industry.
In order to meet this challenge, the automotive industry has to take new and different approaches on how to deal with SW, both from a technical and engineering perspective but also from a business perspective. SW content and also cost - in automotive electronics continue to grow almost exponentially, and Tom de Marco's question, "Why is SW so expensive" has become one of the most frequently asked questions in our industry. In this presentation, I will try to highlight some measures on how to limit SW cost but also try to redirect the discussion towards the role of SW as major added-value and not only major added cost.

• Manfred Broy, TU Munich, Germany
• Ken Butts, Toyota Technical Center, Ann Arbor, MI, USA
• Paul Caspi, VERIMAG, France
• Rance Cleaveland, Univ. of Maryland, Fraunhofer USA Center for Experimental SW Engineering and Reactive Systems, Inc., USA
• Werner Damm, Oldenburg University, Germany
• Klaus Grimm, Daimler Chrysler Corporate Research, Berlin, Germany
• Hermann Kopetz, TU Vienna, Austria
• Pieter J. Mosterman, The MathWorks, Inc., Natick, MA, USA
• Wolfgang Pree, Salzburg University, Austria
• Alberto Sangiovanni-Vincentelli, University of Berkeley, USA
• Bran Selic, IBM SW Group - Rational Software, Kanata, ON, Canada
• Clemens Szyperski, Microsoft Research, USA
• and the organizers

• Alexander Pretschner, ETH Zürich, Switzerland
• Martin Rappl (chair), Technische Universität München, Germany
• Christian Salzmann, BMW Car IT, München, Germany
• Thomas Stauner, BMW AG, München, Germany

1. Markus Bechter, Michael Blum, Hendrik Dettmering and Benno Stützel:
   Compatibility Models in UML
   With the possibility to configure new cars completely individual almost no vehicle of the premium class exists twice. As a consequence the cars differ not only in design but do have different functional features as well. These functionalities are mostly realised by an interaction of software and hardware. Therefore the vehicle manufacturer has to control the crosslinking of such functionalities in the vehicle. The solution of this problem demands dealing with the topic of compatibility of software components. This paper shows how to furnish proof of the compatibility of a variety of software functions with minimal additional modelling and simultaneously reducing time and effort usually needed for the test-based process.
   slides
2. Eckard Bringmann and Andreas Kraemer:
   Systematic Testing of the Continuous Behavior of Automotive Systems
   In this paper, we introduce a new test method that enables the systematic definition of executable test cases for testing the continuous behavior of automotive embedded systems. This method is based on a graphical notation for test cases that is not only easy to understand but also powerful enough to express very complex, fully automated tests as well as reactive tests. This new approach is already in use in several production-vehicle development projects at DaimlerChrysler and at some suppliers.
   slides
3. Pascal Montag, Steffen Görzig and Paul Levi:
   Applying Static Timing Analysis to Component Architectures
   The increase in software functions and software complexity of automotive applications requires appropriate software architectures. A promising approach is the component architecture which also stands in the centre of the automotive standardisation project AUTOSAR[2]. As every embedded real-time system inevitably has upper time bounds, we present an integrated method of timing estimation for highly flexible and variant applications based on a prototype component architecture. Therefore, we develop methods for parameterised timing estimation which depend on the grade of complexity, variability and necessary exactness. The feasibility of the introduced concept is shown in the prototype architecture and a prototype application.
   slides
4. Ina Schieferdecker, Eckard Bringmann and Jürgen Großmann:
   Continuous TTCN-3: Testing of Embedded Control Systems
   The systematic testing approaches developed within the telecommunication domain for conformance and interoperability testing of communication protocols have been extended and broadened to allow the testing of local and distributed, reactive and proactive systems in further domains such as Internet, IT, control systems in automotive, railways, avionics and alike. With the application of these testing principles it became apparent that the testing of systems with continuous systems is different to that of discrete systems. Although every continuous signal can be discretized by sampling methods (and hence mapped to the discrete signal paradigm), abstraction and performance issues in this setting become critical. This paper revises the initial design of Continuous TTCN-3. It presents the concepts for specifying continuous and hybrid test behaviour. The TTCN-3 extensions are demonstrated for a case study.
   slides
5. Arnd Schnieders:
   Modeling and Implementing Variability in State Machine Based Process Family Architectures for Automotive Systems
   In this paper we present an approach for modeling UML State Machine based process family architectures, which is part of our work to transfer Product Family Engineering techniques to process oriented software like automotive systems. For modeling UML State Machine based architectures of automotive system families we introduce a set of variability mechanisms. These allow for taking the different properties of variability realization techniques into account during the development of a process family architecture. We also touch some variability implementation issues in C/C++ and provide a lightweight UML extension for the introduction of variability mechanisms into State Machines.
   slides
6. Ingo Stürmer, Mirko Conrad, Ines Fey and Heiko Dörr:
   Experiences with Model and Autocode Reviews in Model-based Software Development
   Through the introduction of the model-based development paradigm models became first class citizens in the development of in-vehicle software and are thus also object to strict quality assurance. Just as code reviews are widespread in classical software development, models also have to undergo a stringent review procedure particularly if they serve as a basis for automatic software implementation by means of model-based code generators. In addition to model reviews, the generated production code (autocode) must be reviewed by performing so-called autocode reviews. This paper presents our procedure for a combined model and autocode review and provides examples, lessons learned, as well as significant experimental results drawn from a typical automotive embedded software development project.
   slides
7. Shige Wang, Sushil Birla and Sandeep Neema:
   Language for modeling vehicle motion control bahvioral specification
   Error-free engineering of critical, high-integrity applications such as vehicle motion control (VMC) requires unambiguous behavioral specifications. As system engineering progresses from requirements modeling to functional design, to system decomposition and synthesis, to allocation of components on to a distributed platform, the models with these system artifacts must be transferred across different engineering environments and stages without loss of semantics. Modeling environments currently available for industrial use, with their built-in modeling languages, do not provide this capability. In this paper, we identify certain fundamental requirements and show that these requirements are not satisfied in two commercial modeling tools, and current and proposed international standards. A modeling language for VMC application behavioral specifications is proposed as a candidate for standardization and adoption by the community of users in this domain. The language is based on an extended finite state machine paradigm (eFSM) and meets all identified VMC behavioral specification requirements. While its fundamental modeling elements adopt the notation of the Meta Object Facility (MoF) of the Object Management Group (OMG), the language is specialized from a general mathematical modeling framework and adds unambiguous semantics essential to the VMC domain. An experimental prototype of the eFSM has been implemented and evaluated relative to the requirements for modeling high integrity systems. It has also shown that the method for deriving specialized languages from a general mathematical framework for modeling various sub-domains.
8. Kentaro Yoshimura, Dharmalingam Ganesan and Dirk Muthig:
   Assessing Merge Potential of Existing Engine Control Systems into a Product Line
   Engine Control Systems (ECS) for automobiles have many variants for many manufactures and several markets. To improve their development efficiency, exploiting ECS commonalities and predicting their variability are mandatory. The concept of software product line engineering meets this ECS business background. However, we should carefully investigate the expected technical, economical, and organizational effects of introducing the strategy into existing products. Thereafter, a strategy will be derived systematically and realize the desired benefits. This paper reports an experience with the up-front investigation performed for Hitachi’s ECS. We focus on the approach to plan the migration of the existing family of individual systems into a future product line. The approach assesses potential ways of merging software from existing variants and eventually defines a procedure for performing the migration. To get a high quality strategy, we integrate the approach of software measurement, the expertise of software architects, and reverse engineering techniques.
   slides