|
COP6: Validation framework for metrological software
Beneficiaries: Industry, NMIs, calibration service, testing laboratories Task 1: Development of a common understanding of validation of metrological software Aim: clarification of what is expected when metrological software is called "validated", and what must be done when performing a validation Subtask 1.1: Collecting and specifying the issues Subtask 1.2: Classification of different types of validation to meet different requirements (e.g. model validation, validation of numerical stability and correctness, validation of correct implementation)
Task 2: Evaluation of validations performed / further examples that will be performed Aim: Learning from practice and identification of examples worth being followed Subtask 2.1: Sharing information on cases studies already performed or that will be initiated by this COP, and on tools/methods used Subtask 2.2: Evaluation of case studies including feedback for classification and development of criteria Subtask 2.3: Identification of examples to be used as best practice recommendations Subtask 2.4: Initiating new example tools using existing methodologies (e.g. developing new reference datasets for numerical correctness testing)
Task 3: Classification of requirements / user needs Aim: To classify requirements and user needs in order to develop a framework that helps to exploit more systematically the existing knowledge and integration of new developments Subtask 3.1: Development of class definitions and limitations (e.g., for the above mentioned classes) Subtask 3.2: Selection of classes so that validation procedures for practical applications can be provided, and/or validation is urgently required Subtask 3.3: Commenting on how well existing guidance material and procedures meet these requirements and user needs, particularly identifying gaps that need to be filled
Task 4: Development of criteria Aim: Summarising the results as far as possible for recommendations with respect to each of the topics covered in the following subtasks. Subtask 4.1: Classification of software Subtask 4.2: Applicability of existing standards, guides, etc.; agreement on which guidance material can be recommended for use Subtask 4.3: Application of further procedures and methods Subtask 4.4: Acceptance criteria of validations performed Subtask 4.5: Transferability of validations performed for similar but not identical software Subtask 4.6: How to fill the identified gaps in existing guidance material and procedures in order to meet the full set of requirements and user needs
| 
