Critical Software was hired by AgustaWestland to update DGS, an application used by the Royal Danish Air Force to monitor the health and usage of its EH-101 Merlin helicopter fleet. The purpose of this update is to make the aplication work with their new maintenance management system. This project consists of developing the library responsible for the communication with the new system usingWeb Services, called SOAP Adaptor, also supporting the use of a local database, without changing the way DGS works. Also included in the project is an application, SOAP Client, whose purpose is to test the communication library and to create the local database, populating it with data retrieved from the maintenance management system.

The project was developed on top of the .NET platform, in C#, using Microsoft Visual C# 2008. Due to its interface designer, this greatly simplified the development of SOAP Client's graphical interface. The project's outputs were tested using the acceptance tests specified for the previous version of DGS, ensuring its functionality remains unchanged. Since the SOAP Client was more of a support tool than a formal project deliverable, it was tested informally during the project's development.

The project developed at a good pace in spite of a minor delay entering the test phase. This, however, did not affect the final deadline. It was concluded that a Service-Oriented Architecture is a good way to achieve systems integration and that Web Services are a good technology to implement this type of architecture. This is mainly because Web Services do not bring any new technologies into the arena, using instead existing and proven technologies to achieve a new goal. In spite of this, Web Services are not the definitive solution for integration problems. This was demonstrated by the fact that some changes were necessary for the developed library to be able to interact with the client's Web Service, which was supported by BEA WebLogic, a J2EE application server. The SOAP Client application turned out to be extremely useful to test for flaws in the developed code as well as in the Web Service itself. Using only DGS, these flaws would have been very difficult to track down.

In terms of future work, there are two possible approaches. One is to experiment with the use of REST instead of SOAP as the Web Service's message exchange protocol. REST looks promissing especially because of its performance gains and the simplicity of the code required. The other possibility is redesigning the local database's schema. By eliminating some data redundancy, a number of operations would be simplified, with a likely improvement in performance.

 
Top