Observation-based Development of Software Process Baselines: An Experience Report

TL;DR

This paper presents an observation-based approach to quickly develop and baseline software processes for new domains like wireless Internet services, using pilot projects and various modeling and measurement techniques.

Contribution

It introduces a systematic, observation-driven method for creating initial process models and baselines in new software development domains, facilitating rapid process stabilization and understanding.

Findings

Process baselines were quickly established from pilot projects.

The approach improved understanding of process effects quantitatively and qualitatively.

Stable, context-oriented development processes were achieved.

Abstract

The creation and deployment of software development processes for new domains (such as wireless Internet services) is a challenging task due to the lack of knowledge about adequate engineering techniques and their effects. In addition, time-to-market pressure prevents applying long-lasting maturation of processes. Nevertheless, developing software of a predetermined quality in a predictable fashion can only be achieved with systematic development processes and the use of engineering principles. A descriptive approach promises to quickly create initial valuable process models and quantitative baselines that can be seen as starting points for continuous improvement activities. This paper describes the creation of software development processes for the development of wireless Internet services based on the observation of pilot projects that were performed at distributed development sites. Different techniques and tools such as descriptive process modeling, process documentation generation, goal-oriented measurement, and capturing qualitative experience were combined to gain process baselines. Results show that the observation-based approach helped to quickly come up with stable context-oriented development processes and get a better understanding of their effects with respect to quantitative and qualitative means.