RT doctoral thesis
T1 Técnicas de prueba avanzadas para la generación de casos de prueba
A1 Valle Gómez, Kevin Jesús
A2 Ingeniería Informática
K1 lenguajes de programación
K1 Software
K1 téncnicas de prueba
K1 Software testing
AB Software testing is a crucial phase in software development, particularly in contexts suchas critical systems, where even minor errors can have severe consequences. The advent ofIndustry 4.0 brings new challenges, with software present in almost all industrial systems.Overcoming technical limitations, as well as limited development times and budgets, isa major challenge that software testing faces nowadays. Such limitations can result ininsufficient attention being paid to it.The Bay of Cadiz’s industrial sector is known for its world-leading technological projects,with facilities and staff fully committed to innovation. The close relationship betweenthese companies and the University of Cadiz allows for a constant exchange betweenindustry and academia.This PhD thesis aims to identify the most important elements of software testing inIndustry 4.0, based on close industrial experience and the latest state-of-the-art work.This allows us to break down the software testing process in a context where large teamswork on large-scale, changing projects with numerous dependencies. It also allows us toestimate the percentage benefit that a solution could provide to test engineers throughoutthe process.Our results indicate a need for non-commercial, flexible, and adaptable solutions forthe automation of software testing, capable of meeting the constantly changing needs ofindustry projects. This work provides a comprehensive study on the industry’s needs andmotivates the development of two new solutions using state-of-the-art technologies, whichare rarely present in industrial work. These results include a tool, ASkeleTon, whichimplements a procedure for generating test harnesses based on the Abstract Syntax Tree(AST) and a study examining the ability of the Dynamic Symbolic Execution (DSE)testing technique to generate test data capable of detecting potential faults in software.This study leads to the creation of a novel family of testing techniques, called mutationinspired symbolic execution (MISE), which combines DSE with mutation testing (MT)to produce test data capable of detecting more potential faults than DSE alone. Thefindings of this work can serve as a reference for future research on software testing inIndustry 4.0.The solutions developed in this PhD thesis are able to automate essential tasks in software testing, resulting in significant potential benefits. These benefits are not only forthe industry, but the creation of the new family of testing techniques also represents apromising line of research for the scientific community, benefiting all software projectsregardless of their field of application.
YR 2023
FD 2023
LK http://hdl.handle.net/10498/29222
UL http://hdl.handle.net/10498/29222
LA eng
DS Repositorio Institucional de la Universidad de Cádiz
RD 09-nov-2026