Software Cockpit for software development
Software Engineering Intelligence as the basis for high quality
A decisive success factor for software producers is the quality of their software. In the research project ValueSoft within the COMET program, the focus is on support for project management as a profound influence on software quality. However, the project targets not technical project management but software project management based on economic criteria. As a source of information for the project manager via indicators such as metrics and knowledge about interrelationships, Software Engineering Intelligence is provided as a cooperation of all three competence areas at SCCH (DBT, KBT and SET).
Overview
Software systems must meet steadily rising demands regarding stability, performance, usability and maintainability. However, economic indicators such as development time and costs need to be considered as well. In practice, all too often decisions are made at strategic and operative levels without the required background information. There are numerous examples, such as deciding which functionality is to be incorporated in a new release, which architecture to select, how high to set the degree of test coverage, or where to implement measures for quality improvement.
Challenge
Software tools that collect, process and avail the required information for management and the project team. Information is extracted from the project history across the various phases of software development, stored centrally in a data warehouse and depicted in a dashboard. As in an automobile or the cockpit of an airplane, the important information are close at hand. This representation form constantly provides an overview of the status of software development and also enables the analysis of interrelationships and a view of details via a drill down to critical factors.
Solution
For better estimation of future developments of a software project, important factors are the processing and analysis of historical project data as well as prognoses. In this research project, defect prediction is being developed for software components. On the basis of product and process metrics, the defect potential of components of a software system is determined during its development. Thereby the investment for quality assurance in large software systems can be controlled and the individual measures such as component tests, reviews or redesigns can be planned efficiently.
BMD Systemhaus GesmbH, a cooperation partner in the project ValueSoft, already uses SCCH’s dashboard for developing its own business software. The dashboard supports investment estimates and release planning. "Economic aspects move to the foreground and apply for the complete team (testers, developers, project leader, management). The status of a project is depicted in a transparent way for the whole team and the economic goals are visualized. The data are available and can be used optimally for planning and prognosis," reports Dr. Knasmüller, head of software development at BMD.