Home / University / Departments / Mathematics and Computer Science / Section Stochastics / Chair of Probability and Statistics / Projects of Alessandro Di Bucchianico



Successful testing processes require excellence in both software testing and management. In order to support well-founded decisions on issues like resource allocation and software release moments, quantitative procedures are indispensable. Since few testing processes have a deterministic course, statistics is very often an appropriate part of such quantitative procedures. Existing tools for software reliability analysis like Casre and Smerfs do not make full use of state-of-the-art statistical methodology or do not conform to best practices in statistics. Thus, these tools cannot fully support sound software reliability analyses. We decided to build a new tool that
  • uses well-documented state-of-the-art algorithms
  • is platform independent
  • encourages to apply best practices from statistics
  • can easily be extended to incorporate new models.

In order to meet these requirements we decided to use Java for the interface and the statistical programming language R for the statistical computations. R is open-source free software maintained by a group of top-level statisticians and is rapidly becoming the standard programming language within the statistical community. Our tool is a joint project of Eindhoven University of Technology and Refis . The tool development was financially supported by a grant of the Dutch Innovation Platform.

Features of this tool include:

  1. numerically stable algorithms for ML estimation
  2. check for existence of ML estimators
  3. statistically sound workflow
  4. graphical methods (including TTT plot) and trend tests for software reliability growth testing
  5. model selection matrix
  6. model selection wizard
  7. supports both exact data as interval data
Trend tests Model selection wizard Analysis results


Available background information






Last modified: March 7, 2017