2018

Visualization of software bugs

Software bugs occur at different rate depending on the characteristic of the software systems. Automatically collecting bugs from online or offline source (e.g., issue repositories) and visualising them over time help understand the reliability of the system and predict its future behaviour. A simple tool that mines the repository and visualise optimal curves representing the occurrences of bugs over time will be developed.

Input: A dataset of bugs with their occurrence time stamps or an online bug repository (e.g., Jira).

Output: A Java - R  application that connects to the repository or the dataset and visualises both bugs over time and the optimal curve representing their cumulative number.

 

Unit Testing framework for Programmable Logic Controllers

Programmable Logic Controllers are digital systems for process automation. They are typically little or poorly tested. A library for Unit testing and compliant to IEC 61131-3 standards languages with which PLC are developed has been developed. A multiple case study on the efficiency and enhancement of the library for full testing automation will be performed.

Input: PLC developed with IEC 61131-3 compliant languages (e.g., CodeSys)

Output: Enhanced Unit testing library portable on different hardware devices. A study on the portability and efficiency of the tests.

 

Automated UI tests

Testing UI often requires a long manual selection of the input values. Selecting the input values with which to design test with high code coverage is an art. Techniques for UI testing vary depending on the data structures used by the interface. Testing for all combinations of input values is often unfeasible or expensive. Based on the available knowledge of the input space, an algorithm to reduce the combination of the input values will be designed.

Input: UML class diagrams and the Java implementation of such classes and their input space.

Output: An algorithm that identifies combinations of input values with which to design test cases with high code coverage.