SVETLANA will use automated flight data processing tools based on advanced algorithms for knowledge extraction from the data. These algorithms will be designed to analyze data and will be able to identify fault signature and to predict abnormal behaviour. In addition, SVETLANA will introduce a proposal for standards in support of the analysis process. By introducing these results significant gains are expected in the field of flight safety but also in field of maintenance-related data processing.
The SVETLANA concept is based on a two-steps methodology: a highly automated flight data analysis phase, followed by a short expert assessment when abnormalities are detected. The first phase of analysis will use intelligent algorithms with selflearning capabilities, able to analyse all available flight data and at a deeper level. SVETLANA will then allow significantly more flight data to be processed with same or less human effort.
By providing a proposal for the introduction of standards, the implementation of SVETLANA will ease flight data analysis and enable a comparison of results between peer groups of operators. Sharing of such results and knowledge is impossible today with too different analysis methods.
In addition, SVETLANA will be designed to include a feedback loop that aims at detecting future events more efficiently. By introducing a close loop approach, the FDM process will be updated on the bases of previous findings from the flight data analysis. By closing the cycle it is assured that the FDM process is up to date to cope with new emerging changes in operations or safety.
The SVETLANA data mining processes can be applied at a higher frequency compared to classical FDM processes used today. SVETLANA aims at reducing the FDM process to a single phase of automated analysis followed by a human expertise assessment in case of abnormal events. This will shorten the time before feedback to relevant stakeholders is provided.
This process will allow the introduction of other programmes to improve flight operations on the basis of flight data analysis. The potential areas where advanced flight data analysis might bring economical viable improvements are expected to be e.g. fuel consumption optimisation, training programme update, training programme effectiveness assessment, etc…
The concept of SVETLANA will be designed to be adaptable to the existing FDM systems (like Sagem’s AGS) by providing a modular architecture to introduce the new flight data processing capabilities. With this approach, SVETLANA technology can be implemented with a minimum of additional investments in training and integration. The SVETLANA innovation aims at providing new standards to the FDM process, with a complete and systematic data processing allowing 100 % analysis of all flight data.