We are pleased to announce the creation of a new expert network on Model-Driven Engineering. With initial funding from the Engineering and Physical Sciences Research Council in the United Kingdom, MDENet brings together research and practice, in the UK and internationally. We plan to:
- Generate new, high impact research collaborations and projects driven by the software industry or by stakeholders from other subject areas that could benefit from MDE (e.g., artificial intelligence, computational biology);
- Produce and curate comprehensive MDE training material and deliver it through a range of training activities; and
- Increase awareness, and thus impact, of MDE research and technology in the software industry.
If you are interested in finding out more about the network, please register your interest at https://www.mde-network.org/.
The demand for more, and more complex, software far outstrips our ability to ‘produce’ highly-trained software-engineering specialists capable of developing this software. One estimate is that a further 1.6 million ICT professional jobs will need to be filled in Europe by 2030. We need to massively increase the number of people directly involved in software development, but we cannot expect to do this purely by increasing the number of software specialists. In the long term, the only viable approach is to empower more non-computing specialists—that is, more stakeholders with expert domain knowledge but limited or no software-engineering expertise—to leverage the work of software engineers and become active contributors to the creation of the software they use.
Model-driven software engineering (MDE) aims to address this crisis by permitting better separation of concerns. For example, software specialists may develop domain-specific languages (DSLs), which domain experts use to express a model of each concern in terms familiar to the domain experts. To automatically combine concerns and reconcile them as necessary, software specialists develop model transformations. As a result, at least some decisions about the software’s behaviour can be taken directly by domain experts who are not software specialists.
However, in order to realise this vision in full and make its benefits available beyond the group of industrial pioneers who have already been taking up, and reaping the benefits of, MDE, we need to:
- Connect islands of research in academia and industry. MDE research has been driven by individual research groups and industry stakeholders, not always with reference to a wider, shared research vision (though there has recently been some work in this direction). We will run a series of research-oriented sandpit workshops, and manage a seed-corn fund for novel, collaborative and co-creational research, in order to enhance industry–academia collaboration so that future research continues to be driven by real needs.
- Improve training / education materials and opportunities. A successful MDE project requires choosing and applying appropriate MDE techniques and tools. Many current tools are fragile, immature and poorly documented. While there are some curated sources of information (e.g., about UML tools or essential reading), there is a lack of good training materials and there are too few trained graduates. This makes exploring MDE opportunities risky for IT managers, particularly in SMEs. We will build a shared repository for training material (both curating existing materials and developing new materials); run training events; and contribute to MDE curriculum development.
- Increase visibility of MDE and the MDE expertise and successes available. MDE comes with an upfront cost. In today’s agile developer community, the term ‘modelling’ carries negative connotations. At the same time, there is limited general awareness of MDE success stories and the substantive productivity contributions it can offer. The MDE research and user community has traditionally not had a unified presence, instead progress has mostly been made through individual projects. Of course, there are already great places like modeling-languages.com and growing communities like d.strumenta.community. But there isn’t really a central place, targeting people who don’t yet know MDE (as well as people who do), which can increase external visibility of MDE and encourage uptake by providing a central source of curated information on MDE research and achievements, directory of experts, coherent training materials etc. Such a central platform would increase the visibility of opportunities for the economy to benefit from MDE research–industry collaborations. This network, MDENet, will provide it.
MDE clearly has the potential, as demonstrated by the successes already achieved, to fulfil the long-term vision of empowering non-computing stakeholders to co-develop software, bringing people into the heart of software engineering. For it to do so successfully requires sustained and joined-up action from the community as a whole. MDENet will support this. Join us at https://www.mde-network.org/!
I’m a senior lecturer in computer science at King’s College, London. My research focuses on model-driven development, specifically modularity and reuse, graph transformations, search-based approaches.