ADR.OPS.A.035 Data validation and verification

VALIDATION AND VERIFICATION

(a)The processes implemented to carry out validation and verifications should define the means used to:

(1)verify received data and confirm that the data has been received without corruption;

(2)preserve data quality and ensure that stored data is protected from corruption; and

(3)confirm that originated data has not been corrupted prior to being stored.

(b)Those processes should define the:

(1)actions to be taken when data fails a verification or validation check; and

(2)tools required for the verification and validation process.

VALIDATION AND VERIFICATION — GENERAL

(a)Validation

(1)Validation is the activity where a data element is checked as having a value that is fully applicable to the identity ascribed to the data element, or where a set of data elements are checked as being acceptable for their intended use.

(2)The application of validation techniques considers the entire aeronautical data chain. This includes the validation performed by prior data chain participants and any requirements levied on the data supplier.

(3)Examples of validation techniques include:

(i)Validation by application One method of validation is to apply data under test conditions. In certain cases, this may not be practical. Validation by application is considered to be the most effective form of validation. For example, flight inspection of final approach segment data prior to publication can be used to ensure that the published data is acceptable.

(ii)Logical consistency Logical consistency validates by comparing two different data sets or elements and identifying inconsistencies between values based on operative rules (e.g. business rules).

(iii)Semantic consistency Semantic consistency validates by comparing data to an expected value or range of values for the data characteristics.

(iv)Validation by sampling Validation by sampling evaluates a representative sample of data and applies statistical analysis to determine the confidence in the data quality.

(b)Verification

(1)Verification is a process for checking the integrity of a data element whereby the data element is compared to another source, either from a different process or from a different point in the same process. While verification cannot ensure that the data is correct, it can be effective to ensure that the data has not been corrupted by the data process.

(2)The application of verification techniques considers only the portion of the aeronautical data chain controlled by the organisation. Yet, verification techniques may be applied at multiple phases of the data processing chain.

(3)Examples of verification techniques include:

(i)Feedback Feedback testing is the comparison between the output and input state of a data set.

(ii)Independent redundancy Independent redundancy testing involves processing the same data through two or more independent processes and comparing the data output of each process.

(iii)Update comparison Updated data can be compared to its previous version. This comparison can identify all data elements that have changed. The list of changed elements can then be compared to a similar list generated by the supplier. A problem can be detected if an element is identified as changed on one list and not on the other.

VALIDATION AND VERIFICATION TECHNIQUES Validation and verification techniques are employed throughout the data processing chain to ensure that the data meets the associated data quality requirements. More explanatory material may be found in Appendix C (Guidance on compliance with data processing requirements) to EUROCAE ED76A ‘Standards for Processing Aeronautical Data’.