Formalizing standards and regulations variability in longlife projects. A challenge for Model-driven engineering

by Nicolas Sannier, Benoit Baudry, Thuy Nguyen
Abstract:
Safety regulations and standards imposed by national regulators on nuclear power plant Instrumentation & Control (I&C) systems provide high-level requirements, recommendations and/or guidance expressed in natural language. In many cases, this leaves a large margin for interpretation, not all of which are acceptable to a given regulator. The acceptable/accepted interpretations are currently not systematically and formally documented, which is an issue when a licensee faces a given regulator for the first time and does not know whether their practical approaches and solutions, which might have met with regulatory approval elsewhere, will be endorsed by this regulator. Model-driven engineering techniques should allow explicit recording of the interpretations that have been accepted past projects and provide analysis capabilities to follow their evolution or stability. It should also help to ensure that one design complies with the regulatory requirements and practices of the targeted countries. As an initial analysis in this direction, this paper proposes a first glance on variability dimensions that should be captured to manage standards and regulatory requirements and their practical interpretations. We then review existing approaches for requirements relations modeling and identify gaps between these solutions and the problem we want to address for systems requirements management.
Reference:
Formalizing standards and regulations variability in longlife projects. A challenge for Model-driven engineering (Nicolas Sannier, Benoit Baudry, Thuy Nguyen), In Proceedings of the Model-Driven Requirements Engineering workshop (MoDRE’11) at RE’11, 2011.
Bibtex Entry:
@inproceedings{Sannier11,
	Abstract = {Safety regulations and standards imposed by national
        regulators on nuclear power plant Instrumentation &
        Control (I&C) systems provide high-level requirements,
        recommendations and/or guidance expressed in natural
        language. In many cases, this leaves a large margin for
        interpretation, not all of which are acceptable to a given
        regulator. The acceptable/accepted interpretations are
        currently not systematically and formally documented, which
        is an issue when a licensee faces a given regulator for the
        first time and does not know whether their practical
        approaches and solutions, which might have met with
        regulatory approval elsewhere, will be endorsed by this
        regulator.
        Model-driven engineering techniques should allow explicit
        recording of the interpretations that have been accepted past
        projects and provide analysis capabilities to follow their
        evolution or stability. It should also help to ensure that one
        design complies with the regulatory requirements and
        practices of the targeted countries. As an initial analysis in
        this direction, this paper proposes a first glance on
        variability dimensions that should be captured to manage
        standards and regulatory requirements and their practical
        interpretations. We then review existing approaches for
        requirements relations modeling and identify gaps between
        these solutions and the problem we want to address for
        systems requirements management.},
	Address = {Trento, Italy},
	keywords = {RE, MDE},
	Author = {Sannier, Nicolas and Baudry, Benoit and Nguyen, Thuy},
	Booktitle = {Proceedings of the Model-Driven Requirements Engineering workshop (MoDRE'11) at RE'11},
	Month = aug,
	Title = {Formalizing standards and regulations variability in longlife projects. A challenge for Model-driven engineering},
	X-Country = {IT},
	X-International-Audience = {yes},
	X-Language = {EN},
	X-Proceedings = {yes},
	x-abbrv = {MoDRE},
	Year = {2011}}