Multisite Software Engineering

Current system development methodologies do not address the issue of collaborative multi-site distributed software development environments. Most of the software process models are based on the assumption that the team undertaking the software development is based at a single central site and that the project and user managers have ready access to the software development team. In practice, this assumption is increasingly not fulfilled.

Current system developments, based on dynamic business models, use a variety of technologies, a range of modelling and methodological skills, rely on complex project management skills, and require increasing domain knowledge. In most single sites, we cannot find a coherent collection of people with skills in all these areas. To successfully deliver a large complex information system and to reduce the cost of software development, we often outsource to obtain the skills needed. This means that specialised groups will have to work together remotely to achieve a common integrated development goal.

One of the most significant issues is the need to establish a common underlying structure that provides communication, interaction and management between the different parties engaged in the software development, as well as overcoming differences in terminology, opinion, expertise and understanding of the software development project. In our project, we propose the use of an ontology driven architecture to provide a framework in which multi-site software development can operate.

In general, the problem areas to be explored in this project can be divided into the following research issues:

Define the concepts, requirements, and representation of an ontology for multi-site distributed software development.

1. Define an ontology-based software development architecture which addresses the needs of the collaborative, multi-site and distributed environment. This will address issues such as awareness, access control, security, communication and group decision support.
   
2. Can one carry out the definition of permission-controlled interfaces of objects and components in a form suitable for multi-site development?
   
3. Development of a flexible workflow model that captures asynchronous scheduling of activities at different sites and the development of a project management approach for a distributed ontology-based project architecture.
   
4. Develop a prototype system that facilitates information exchange over the Internet and enforces security mechanisms for data transfer.
   
5. Benchmarking and Field Testing. The Industry Collaborators have all agreed to put existing projects that they are currently working on at multi-sites on this prototype platform to allow field-testing, evaluation and benchmarking against their current practices.
   
   Specifically:
   
   
   1. Azurn International is collaborating with Curtin University to redesign the ways the firm currently manage its multi-site R&D and customer support activities between Australia and the United States. Future plans in the next 12 to 18 months include expanding to other countries, as the company internationalises.
      
   2. C-E Solutions works with projects that are spread between Melbourne and Newcastle.
      
   3. Computer Associates Inc. with projects and customers who are spread all around Australia and overs

Ontology is the term used to refer to a conceptualisation of some domain of interest, which may be used as unifying framework to solve the problems. A key feature of an ontology is agreement about shared conceptualisations. The use of an ontology reduces conceptual and terminological confusion by providing a unifying framework within an organisation or community of users. In this way, an ontology enables shared understanding and communication between the front-end groups, backend groups and the application groups. In other words, different groups producing their own databases and residing at different sites can share a conceptual model.

Using an ontology, the system can construct a normative model. This creates a semantic for the system and an extendible model that can later be refined, and which allows semantic transformations between different contexts from different sites. The ontology provides unambiguous definitions for terms in a software system or database system. Any set of software tools should be able to maintain consistency among themselves and the ontology though they need not be uniform. The ontology provides, through the normative model of the system, integration of different user perspectives. It also allows interoperability among systems by facilitating translating between different databases.