MiSE 13 - San Francisco, USA

Workshop Overview

The purpose of this 2-day workshop is to study and advance the effective use of models in the engineering of software systems. In particular, we are interested in the exchange of experiences and innovative technical ideas related to modeling. Engineers have used models to manage complexity for centuries, and there is a growing body of work on the use of models to manage inherent problem and solution complexity in software development. The use of software models will become more prevalent as methodologies and tools that manipulate them at various levels of abstraction become available. A secondary goal of the workshop is to further promote cross-fertilization between the modeling communities (e.g., MODELS) and software-engineering communities.

We take a broad view of what models are and how they are used, including

• Exploration: Models are used to explore and learn about the problem to be solved, where the “problem” can be, for example, requirements identification, system specification, system or component design, complex protocol or algorithm design.
• Communication: Communication models are used to document software decisions (e.g., requirements, designs, and deployment decisions). Support for downstream activities: We use software models to answer questions or check properties (e.g., correctness, fitness for use) of the modeled artifact, to generate other artifacts, or to configure existing systems.
• Support for downstream activities: We use software models to answer questions or check properties (e.g., correctness, fitness for use) of the modeled artifact, to generate other artifacts, or to configure existing systems.
• Configurability and adaptation: We use models at runtime to configure the system, to adapt it to changed needs of the users. A model of the environment also allows a system to capture its knowledge about the environment it controls or communicates with.

Workshop activities will focus on analyzing successful and unsuccessful applications of software modeling techniques to gain insights into challenging modeling problems, including: (1) identifying, describing, and using appropriate abstractions, (2) supporting incremental, iterative development through the use of appropriate model composition, transformation and other model manipulation operators, and (3) automated analysis of possibly large and possibly incomplete models to determine the presence or absence of desired and undesired properties.

Topics of interest include

• Modeling notations and tools
• Metamodeling
• Abstractions and modeling methodologies
• Model-based analysis and synthesis
• Model transformation and composition
• Model evaluation
• Model management
• Extracting models from software artifacts (e.g., program understanding)
• Use of models for downstream activities (e.g., generating test cases, using models at runtime)
• Modeling the systems environment
• Models at runtime (e.g. for software adaptation)
• Empirical studies
• Domain-specific modeling
• Variability management using models
• Model reuse
• Other uses of modeling

Workshop Program

Saturday May 18

8:30 Session #1: Introduction and Keynote

Session Chair: Joanne Atlee

• Why We Model: Using MBD Effectively in Critical Domains (Slides)
  Michael Whalen, University of Minnesota, USA
  
  Model-Based Development (MBD) is a paradigm in which software and systems development are focused on high-level executable models of the system to be fielded. These models allow a wide range of exploration and analysis including high-fidelity simulation, test-case generation, and even proofs to be performed at low cost early in development cycles. MBD is seeing widespread adoption in a variety of domains, especially in critical systems such as aerospace and medical devices. In this talk, I describe how to use model-based development effectively in the development of critical software, using examples from the aerospace domain. I focus on V&V of models, as this is an important cost driver for critical systems, and discuss benefits of using automated verification techniques for both unit testing to "unit proving". I also describe several possible pitfalls that can occur when adopting model-based development at scale and steps to take to avoid them.

10:30: Break

11:00 Session #2: Model Representation

Session Chair: Jeff Gray

• Issues in Representing Domain-Specific Concerns in Model-Driven Engineering
  Lionel Montrieux, Yijun Yu, Michel Wermelinger and Zhenjiang Hu
• MAV-Vis: a Notation for Model Uncertainty (Slides)
  Michalis Famelis and Stephanie Santosa
• Modes, Features, and State-Based Modeling for Clarity and Flexibility (Slides)
  Anitha Murugesan, Sanjai Rayadurgam and Mats Heimdahl

12:30: Lunch

2:00 Session #3: Applications of Modeling

Session Chair: Robert Baillargeon

• Model Based Control for Multi-Cloud Applications (Slides)
  Marco Miglierina, Giovanni Paolo Gibilisco, Danilo Ardagna and Elisabetta Di Nitto
• Design Module: A Modularity Vision Beyond Code (Not Only Program Code But Also a Design Model Is a Module) (Slides)
  Naoyasu Ubayashi and Yasutaka Kamei
• Highlighting the Challenges of Model-Based Engineering for Spaceflight Software Systems
  Robert Pettit and Navneet Mezcciani

4:00: Break

4:30 Session #4: Poster Session

Sunday May 19

8:30 Session #5: Keynote

Session Chair: Richard Paige

• Research-Based Innovation in Model-Driven Engineering: Project Experience and Lessons Learned (Slides)
  Lionel Briand, University of Luxembourg, Luxembourg
  
  Engineering research needs to be informed by practice to be relevant and have impact, and industrial innovation relies on research to fill the gaps in knowledge and to pave the way for new tools, technologies, and services. With a focus on Model-Driven Engineering (MDE), this talk will report on my experience from a number of recent successful research projects conducted in various industry sectors. I will take a retrospective approach to examine the way I collaborated with the industry partners and elaborate on the decisions that I believe contributed to the effectiveness of the collaborations. I will then summarise the lessons learned from this experience and illustrate these lessons using examples from the projects.

10:30: Break

11:00 Session #6: Model Analysis and Development Support

Session Chair: Marsha Chechik

• Light-weight Analysis of Software Design Models at the Whiteboard (Slides)
  Alfredo Motta, Nicolas Mangano and André van der Hoek
• Complementing Model-Driven Development for the Detection of Software Architecture Erosion (Slides)
  Sebastian Herold and Andreas Rausch
• Enhancing Version Control with Domain-Specific Semantics
  Matthieu Foucault, Sébastien Barbier and David Lugato

12:30: Lunch

2:00 Session #7: Runtime models

Session Chair: Bernhard Rumpe

• A UML Profile for Dynamic Execution Persistence with Monitoring Purposes (Slides)
  Eladio Domínguez Murillo, Beatriz Pérez and María A. Zapata
• Requirements Models for Design- and Runtime: a Position Paper (Slides)
  Alexander Borgida, Fabiano Dalpiaz, Jennifer Horkoff and John Mylopoulos

Session #7b: Short Papers

Session Chair: Bernhard Rumpe

• Model Oriented Programming: Bridging the Code-Model Divide (Slides)
  Omar Badreddin and Timothy C. Lethbridge
• A Model-Based Approach to Language Integration (Slides)
  Federico Tomassetti, Antonio Vetro', Marco Torchiano, Markus Voelter and Bernd Kolb
• Prioritizing Software Anomalies with Software Metrics and Architecture Blueprints: A Controlled Experiment (Slides)
  Everton Guimaraes, Alessandro Garcia, Eduardo Figueiredo and Yuanfang Cai

4:00: Break

4:30 Session #8: Panel
Question: What should be the focus of software modeling research over the next 10 years?

Session Chair: Robert France

• Panelists: Lionel Briand, Connie Heitmeyer, Richard Paige, Michael Whalen
• Panel Appetizer:
  • A question that is often asked of researchers in the software modeling community is: "What is the problem addressed by researchers in the software modeling community?". The answer is typically a variant of the following: “To significantly reduce the time, cost, and effort required to develop complex software-intensive system that meet stringent quality requirements through the use of models that are fit-for-purpose”. This is still the focus of many research programs, but we’re also witnessing a broadening of the target problems to include not only development problems, but problems that occur throughout the lifetime of a complex system; from conception to retirement. For example, there is emerging work on the use of models to manage software at runtime (sometimes referred to as models@run.time). The concern space targeted by software modeling researchers is thus evolving to encompass a wider range of problems that occur in the lifecycle of complex systems, as it should. One can envisage a time when models become the primary means for managing work throughout a software system’s lifetime.
  • This is not to say that development concerns (e.g., requirements, architecture, detailed design concerns) no longer presents challenges to software modeling researchers. It is evident, that new and emerging types of software systems present good opportunities for the use of modeling techniques, and, at the same time, pose significant challenges to the software modeling community. For example, developing systems that integrate mechanical, software and electrical engineering subsystems (e.g., cyberphysical systems) is currently very challenging. Developers in the mechanical, electrical and software engineering spaces rely on different theories and use different development techniques and processes. The challenge is to develop good bridges across these theories and practices. Cloud systems will soon be integrated with non-cloud-based systems, e.g. for storage and handling of masses of sensor data of various sources (car, home, body, factory, energy, city, state and governance), and will have an impact on systems that connect to the Internet (“Internet of things”). There is thus a need to investigate ways in which cloud-based concerns can be addressed in models.
  • In this panel, we ask the panelists to reflect on the current state of software modeling research and identify some of the major challenges that should be the focus of research over the next 10 years. Panelists can choose to delve into 1 challenge or to give an overview of 3 to 5 challenges.

Important Dates

Workshop Format

The workshop will be highly interactive and focused on the sharing of ideas and on progressing towards a shared vision of research goals. Included in the program will be paper presentations, open discussions, at least one panel discussion, posters, and a keynote talk. Paper presentations will be grouped into sessions of three 20-minute presentations, with 30 minutes of discussion at the end of each session. Our goal is not to create a mini-conference and all attendees will also have an opportunity to discuss topics emerging from the collective presentations.

Submission Guidelines

Prospective participants are invited to submit a 4-6 page position or technical paper. Workshop papers must follow the ICSE 2013 Format and Submission Guideline. All submissions will be reviewed by members of the program committee and the organizing committee for quality and relevance. Accepted papers will become part of the workshop proceedings and will be entered into the ACM and IEEE Digital Libraries.

EasyChair will be used to manage the submission review process. The submission site is here: http://www.easychair.org/conferences/?conf=mise2013

Organizing Committee

• Joanne M. Atlee (primary contact), University of Waterloo, Ontario, Canada
• Robert Baillargeon, Sodius, East Amherst, NY, USA
• Marsha Chechik, University of Toronto, ON, Canada
• Robert France, Colorado State University, USA
• Jeff Gray, University of Alabama, USA
• Richard Paige, University of York, UK
• Bernhard Rumpe, RWTH Aachen University, Germany

Program Committee

• Lionel Briand, University of Luxembourg, Luxembourg
• Manfred Broy, Technical University of Munich, Germany
• Krzysztof Czarnecki, University of Waterloo, Canada
• Juergen Dingel, Queen's University, Canada
• Geri Georg, Colorado State University, USA
• Mats Heimdahl, University of Minnesota, USA
• Michael Jackson, The Open University, UK
• Jean-Marc Jezequel, INRIA & Univ. Rennes 1, France
• Dimitris Kolovos, University of York, UK
• Jeff Kramer, Imperial College London, UK
• Ana Moreira, Universidade Nova Lisboa, Portugal
• Alfonso Pierantonio, University of L’Aquila, Italy
• Awais Rashid, Lancaster University, UK
• Andrey Sadovykh, Softeam, France
• Sebastian Uchitel, Universidad de Buenos Aires, Argentina
• Steffen Zschaler, Kings College, UK

Older Mise Workshops

• older MISE instances