Data migration within aviation is something which is not frequently done within airlines. Mostly is happens once every 10 to 15 years when important systems, such as MRO / M&E software, or Flight Planning systems are replaced. This kind of projects are rarely a one-way trip from system A to system B. Instead, a project of any complexity transforms data substantially and involves multiple data sources to be transformed into a new target. Based on years of experience we have listed some tips:
#1: It's never about the data migration itself
Data Migration is never the purpose on its own. It always coincides with a larger project often involving replacing critical airline system and introducing new working procedures and processes within the airline. This means that data migration is the enabler of a new system being introduced within the airline and/or new working procedures and processes being introduced. This makes that change management is equally important within data migration projects as the migration of the data itself.
#2: History will unveil itself
Migration of data will also mean that historic data will get scrutinized. Particularly when migrating Aircraft Airworthiness data this can lead to some previously unknown surprises to be unveiled. Including maintenance task overruns or wrongly installed aircraft components. Another example includes spare parts inventory, all too often a data migration project will point out the fact that old systems indicated certain spare parts to be stock despite these not being in stock. This could result in significant inventory write-offs. It’s important to be transparent about this to all stakeholders, even local CAA’s. After all, there is a reason you have started this project.
Before getting started with your data migration project, outline the complete project and processes in a detailed plan. This project plan should also include the names of people that are responsible for certain aspects of the project. These people should have the decision-making power to avoid unnecessary long discussions and to drive decision making when required. Even more important, stick to the developed plan if you want to stay on time and budget. Working procedures and processes need to be clear and knowledge of the new MRO software among the key-user should develop continuously. This is necessary in order to understand the possibilities the new MRO system is offering and to map the current data accordingly. By following this approach it’s a linear process and pressure on the project is steady (Figure 1). Otherwise it can happen that pressure is rising towards the end of the project (Figure 2) and by all means this should be avoided.
Figure 1: Linear Curve
Figure 2: Exceptional Curve
#4: Data sources
In the beginning make an overview of all data sources to avoid overseeing anything in a later stage and ensure that you can access all sources during the whole project.
Currently planning or conducting a data migration project and need support or advise?
Schedule a free consultation session with one of our consultants to discuss your needs and problems.
#5: An iterative ETL process is not optional
Many of the best practices of data migration centre on its iterative nature. It involves mapping arcane legacy data, modelling targets, developing transformations, and creating test datasets/ environments. The process is iterative, too, involving multiple rollout and handoff phases that may span across several months. Managing such as process without automated ETL (extract-transform-load) tools is equal to setting up the project to fail.
#6: User acceptance tests
Schedule several user acceptance tests before go-live, to see if everything is fit for go-live as the user acceptance tests are detailed tests of the data and processes. If the user acceptance tests are passed successfully, all is well prepared for go-live and unnecessary surprises can be eliminated to a minimum.
#7: People are as important as data
As mentioned earlier, data migration projects rarely stand on their own. They are normally part of a larger company initiative of replacing airline critical systems and adopting new working processes. This makes that the people are equally important as the data itself. Typically, you would need people involved that can handle pressure, are able to distinct important issues from minor ones and know how to get things done. Obviously getting people involved that have done migration projects previously is a huge advantage.
#8: Cooperation is the key
As data migration is not a daily activity within airlines, all too often these projects get outsourced to suppliers which do have this experience. As important as the track record and know-how of the selected supplier, are the people from the airline that get involved in the data migration project, especially the key-users. It’s the staff of the airline that knows the fine details and peculiarities of the past as well as the internal politics and informal organization. Next, the key-suers work with the data on a daily basis and understand the meaning and value of the data. Among the key-users should also be decision-makers in order to drive decision making when required.
#9: Be open for new things and be prepared for surprises
Last but not least, be open for changes and new ideas. Just because you have always done it this way does not mean that it is the best way. Be prepared for surprises, as you will encounter things in your data that you thought were not possible.