1st Workshop on Pains in Model-Driven Engineering Practice
Co-located with the ACM/IEEE 21th International Conference on Model Driven Engineering Languages and Systems
16 October 2018, Copenhagen, Denmark
Theme and Goals
Model-driven Engineering aims at making software engineering more efficient through application of abstract (domain-specific) models, which are supposed to reduce the conceptual gap between problem domains and running software. Consequently, research in MDE must be driven by industrial requirements and many requirements have been met in the past.
In other words, software engineering research must always be solidly rooted in industrial needs. Such needs must be given a medium to be clearly communicated between industry and research. Otherwise, research becomes untethered while the industry's needs are not met. The objective of the PAINS workshop is to bring together people from industry and the research community to discuss concrete pains, issues and challenges faced in the industrial practice of MDE.
Previous approaches to eliciting MDE requirements often brought forth academic perspectives. With a unique focus on the needs of industrial practitioners, the PAINS workshop aims to close the gap between industrial requirements in model-driven engineering and academic solutions. To achieve this, we propose a workshop format of short presentations of industrial pains, followed by problem-specific or domain-specific breakout sessions better accessible to industrial practitioners.
Discussions at the 2018 Winter Modeling Meeting have shown that there can be gap between industrial requirements ("what is needed") and scientific contributions ("what is wanted"). At PAINS, we aim to provide a forum for industrial requirements elicitation that helps the MODELS 2018 community to reduce this gap.
To elicit current industrial requirements on MDE that allow the community to move forward the adoption of MDE, PAINS presents a forum for industrial and academic practitioners without the barriers of traditional workshops (i.e., mini conferences with rather long papers).
Instead, we call for contributions as (extended) abstracts or presentation slides describing "pains in MDE". Through this, we aim to lower the contribution barrier for industrial practitioners and attract interesting contributions on MDE challenges "in the field", which can include contributions from MDE researchers working with industry as well.
With PAINS, we aim at increasing the number of industrial contributions on challenges in MDE. To this effect, we aim to lower the entry barrier for industrial participants by calling for (extended) abstracts of two pages or presentation slides only. Topics of interest include:
- Pains from applying research prototypes in industrial practice and why these were insufficient
- Pains from missing research solutions to challenges in industrial practice
- Pains in implementation MDE theories in practical tooling
Contributions will be evaluated with respect to their the relevance, grandness and soundness of challenges with respect to the MODELS community. The reviews will also filter challenges that are considered trivially solved. Due to the delicate nature of companies presenting open challenges, PAINS will not have formal proceedings. If the workshop participants, however, are interested we would like to report a summary of challenges and possible approaches to solutions based on the breakout sessions' results afterwards. Ideally, these solutions can be mapped to MSc thesis-sized research proposals aimed to immediately address the elicited pains.
The workshop will be structured into four parts:
- Introduction and motivation followed by invited talks.
- Flash talks of 5-10 minutes each, presenting concrete "MDE pains"
- 2-3 rotating breakout sessions on aggregated challenge topics (e.g., specific domains or modeling techniques)
- Presentation of breakout session results and plenary discussion of the workshop outcomes (e.g., continuation, joint report, further efforts).
Stefan Kriebel (BMW Group): Pains in Modeling: SysML-based Deployment in an Engineering Domain
The continuous main challenge in the development departments of mechanic and electric engineering industries (e.g., automotive) is the well-known decrease of project time and cost and the simultaneous increase of quality. In addition the intra and extra communication requirements of the products are increasing exponentially due to the strong increase of interdependent functionality. Thus customer expectations are often not sufficiently achieved. The never ending continuous improvement story of well-known and formerly approved system engineering processes is definitely told. A disruptive change seems to be necessary to survive the quickly changing market situation, which is strongly influenced by new and quickly changing regulations, new competitors and new customers. (Semi-) formal modelling of requirements and architecture is introduced since the 1990s particularly in the computer science and software domains. The main purpose of these techniques was to overcome the so-called software crisis, i.e. to meet the planned software project objectives in time, budget and quality. As a matter of fact, these techniques made the computer science and software domain becoming reliable and sustainable. In our contribution we firstly want to share some experience of a serious industrial attempt to adapt the (semi-) formal modelling approach to the engineering domain. UML and SysML modelling was applied to the system engineering of the electric powertrain in a real and thus industrial-scale development project. Secondly, the asset of this approach to digital and agile transformation, particularly focusing on development and decision culture is pointed out.
Dr. Stefan Kriebel is assigned Head of Projects Electric Powertrain and responsible for the systems engineering approach and the processes applied at the development department Electric Powertrain at BMW AG, Munich, Germany. In addition he is lecturer at the RWTH Aachen, Germany, at the chair Software Engineering of Prof. Dr. Bernhard Rumpe.
Nikolaus Regnat (Siemens): Applying model based approaches in industry environments: key challenges and issues
Nikolaus Regnat uses the SysML as a prominent example to describe why modeling languages often do not live up to their expectations and why introducing them into an organization frequently fails: Many things can (and often will) go wrong when trying to introduce a model based approach. From the typical dead-end streets regarding how to handle the introduction to having to deal with the management of the organization. There are also language shortcomings and tool issues and the prominent lack of focus on the end users of the modeling language. Languages like SysML were built by modeling experts for modeling experts ignoring the fundamental fact that most end users are not and probably never will be modeling experts.
Nikolaus Regnat is a senior key expert for model based development at Siemens AG. He has been working for the past 15 years on model based development in industry environments and has helped various Siemens business units to adopt model based development methodologies to their organizations. He is also frequently involved in research projects, the latest one being CrESt (Collaborative Embedded Systems) where he is currently leading the sub-project dealing with model based development methodologies.
- Gerald Stieglbauer, Philip Langer and Stefan Sobernig. User Experience for DSLs - Aspects in regard to the Introduction of MDE within an Industrial Environment.
- Jonathan Co, Richard F. Paige, Peter Thomas and Stephen Hobson: The Pains of Working with Spreadsheets in Model-Driven Engineering.
- Arne Nordmann: Pains in Adopting MDE for Safety in Highly-Automated Driving.
- Thibaud Thomas and Sutra Iris: Model Based System Engineering Challenges @ PLASTIC OMNIUM.
- Garry Craig: Code Generation Pains.
- Elena Strabykina and Mattias Mohlin: Simplification and dynamic customization of MDE tools.
- Vinay Kulkarni and Sreedhar Reddy: Pains with MDE.
- Robert Pettit: Pains in Model-Based Engineering: Views from an IV&V Perspective.
- Juha-Pekka Tolvanen: How to provide “experience” to doubters? On introducing MDE into an organization.
- Levi Lucio: The PAINS after 15 years of practicing MDE.
We welcome flash talk proposals in the form of extended abstracts (maximum 2 pages) or slide decks. Submissions must be in PDF format and submitted via Easychair: https://easychair.org/conferences/?conf=pains18
Please copy and paste the following information into the abstract field on the easychair submission.
- Company domain (e.g., aerospace, healthcare)
- Size of company
- Your role
- Years of experience
- Years of experience with MDE (Model-Driven Engineering)
- Describe in 1-2 paragraphs your MDE PAINs
Submission Deadline: July 17, 2018
Notification: July 29, 2018
Workshop Date: October 16, 2018
- Francis Bordeleau (Cmind Inc)
- Michalis Famelis (Université de Montréal)
- Jennifer Horkoff (Chalmers and the University of Gothenburg)
- Andreas Wortmann (RWTH Aachen University)