Comparing and contrasting model-driven engineering at three large companies

Abstract

Hutchinson et al. recently carried out an interview-based study of how model-driven engineering is practiced in 17 companies. Their results are revealing: they found that successful MDE companies develop domain-specific languages; are motivated by a clear business case; and are committed at all levels of the organization. Whilst the results are useful, the study is a very broad one, with one or two interviewees per company. This paper supplements Hutchinson's study by focusing on three large companies, but studying the companies in depth through a number of interviews at each. These three companies have a number of things in common -they are all large companies applying MDE and are all undergoing a parallel transition to agile methods. The paper reflects on similarities and differences in the way these companies apply model-driven engineering, but also discusses how findings from this study either validate or refute those of Hutchinson et al.

Considerations about quality in model-driven engineering

“…(Burden et al, 2014), and which MDE features mesh most easily with features of organizational change? which create most problems?…”

“…The absence of support for MDE tools and the lack of trained people require that great effort be made to adapt to the context of the organization with probably less that optimun results (Burden et al, 2014). This issue leads to problems with the followings: customization, tailoring, and interoperability among modelling tools (Burden et al, 2014) (Mohagheghi et al, 2013b), management of traceability with several tools (Mohagheghi et al, 2013b), the high level of expertise and effort required to develop a MDE tool (Burden et al, 2014) (Mohagheghi et al, 2013b), tool integration (Baker et al, 2005) (Burden et al, 2014) (Mohagheghi and Dehlen, 2008b) (Mohagheghi et al, 2013a), the dissatisfaction of MDE practicioners with the available tools (Tomassetti et al, 2012), the lack of technological maturity of the tools (Mohagheghi et al, 2013a), the scaling of the tools to large system development (Mohagheghi and Dehlen, 2008b), poor user experience (Mohagheghi et al, 2009b), too many dependencies for adopting MDE tools (Whittle et al, 2013), and poor performance (Baker et al, 2005).…”

“…Several reports have pointed out issues in modeldriven adoption that are related to the misalignment between the model-driven principles and the real context (Burden et al, 2014) (Whittle et al, 2013) . Some of these include the overload imposed by the model-driven tools, the lack of traceability mechanisms, and the lack of support for the adoption of model-driven strategies in organizational/development processes.…”

A framework for evaluating the quality of modelling languages in MDE environments A framework for evaluating the quality of modelling languages in MDE environments

“…The absence of support for MDE tools and the lack of trained people require that great effort be made to adapt to the context of the organization with probably less that optimun results (Burden et al, 2014). This issue leads to problems with the followings: customization, tailoring, and interoperability among modelling tools (Burden et al, 2014;Mohagheghi et al, 2013b), management of traceability with several tools (Mohagheghi et al, 2013b), the high level of expertise and effort required to develop a MDE tool (Burden et al, 2014) (Mohagheghi et al, 2013b), tool integration (Burden et al, 2014) (Mohagheghi and Dehlen, 2008b) (Mohagheghi et al, 2013a), the dissatisfaction of MDE practicioners with the available tools (Tomassetti et al, 2012), the lack of technological maturity of the tools (Mohagheghi et al, 2013a), the scaling of the tools to large system development (Mohagheghi and Dehlen, 2008b), poor user experience , too many dependencies for adopting MDE tools , and poor performance .…”

“…Several reports have pointed out issues in model-driven adoption that are related to the misalignment between the model-driven principles and the real context (Burden et al, 2014) . Some of these include the overload imposed by the model-driven tools, the lack of traceability mechanisms, and the lack of support for the adoption of model-driven strategies in organizational/development processes.…”