Project Abstract

Software testing makes up a large portion of the development lifecycle and often times large projects have labour intensive testing processes. This project aims to explore the viability of using evolutionary algorithms to create useful test data for systems that have complex testing requirements. A genetic algorithm was selected to procedurally generate railway map data for Siemens. An iterative development method was chosen to produce an algorithm capable of generating maps with complex and useful features. The prototype software demonstrates a promising future for the technology as it successfully generates interesting results with various tuning parameters which allow the algorithm to produce varied and unique maps. We have shown that genetic algorithms are capable of producing complex test data with multiple simultaneous and unique requirements however quantifying what makes a good or bad solution is a challenge that needs to be explored further.

Keywords: Software Testing, Genetic Algorithm, Procedural Generation

 Conference Details

Session: Poster Session B at Poster Stand 70

Location: Sir Stanley Clarke Auditorium at Wednesday 8th 09:00 – 12:30

Markers: Markus Roggenbach, Tom Owen

Course: BSc Computer Science, 3rd Year

Future Plans: I’m looking for work