I’m currently doing my research within the SQuaD research group. SQuaD is a research group located in the Mathematics and Computer Science department at Karlstad Univesity, Sweden. In general, the group is focusing on Software Quality and Digital Modernisation and aims at developing methods and techniques to continuously preserve, improve, and adapt quality attributes of long-living software systems to allow such systems to be maintained and to evolve more efficiently. My research specifically focuses on developing software testing strategies for modern software systems. One of the goals is to help industrial partners for developing cost-effective and advanced testing techniques for real-world systems. Currently, we are working on six research topics related to software engineering.
- Combinatorial Interaction Testing
- Search-based Software Engineering
- Testing of machine learning systems (also known as machine learning testing)
- Quality aspects of IoT systems
- Smart device app testing (e.g. Mobile app testing and Smart TV app Testing)
- Architectures and process setup for test automation
For collaboration and student supervision (Ph.D. & Master), please contact me bestoun@kau.se
Current Projects
Project Name: AIDA visit the webpage here
AIDA – A Holistic AI-driven Networking and Processing Framework for Industrial IoT
Project Description: The research project AIDA focuses on the interaction of Computer Networking, Distributed Systems, and Software Testing. The project builds on three separate sub-projects that, taken collectively, build a Holistic AI-driven Networking and Processing Framework for InDustriAl IoT. The synergy question is: How to enable trustworthy data-driven real-time industrial IoT applications where time-sensitive networks, low latency computing platforms, and trustworthy AI algorithms play a key role.
I lead the software testing subproject in AIDA and it focuses on building Testing infrastructure for trustworthy data-driven IoT systems. The primary aim of this subproject is to design, develop, and deploy a new testing infrastructure that promotes the trustworthiness of data-driven IoT systems. The testing infrastructure will take into consideration the continuous evolution of the system and also the unpredictable behavior of the ML algorithms. Having such a testing infrastructure will prevent the system from storing noisy and corrupt data, avoid the production of inaccurate or imprecise decision results, and ensure the system’s quality while evolving over time.
Project duration: 2020-2023
Project Name: SmartForge
Smart Forge – Sustainable production through AI controlled forging oven
Project Description: The project´s objective is to create a prototype of an AI-based process control system that with high precision can autonomously control Bharat Forge Kilsta´s forging line in such a way that the steel rod temperature, just before forging, falls within the temperature range that optimizes the properties of the forged component. The direct purpose is to reduce scrap, which today is very costly and thus a major challenge for the company. Expected results are primarily a sharp reduction in scrap, and expected effects include secured Swedish jobs, an increased rate of innovation in Swedish industry, as well as an increased interest in heavy process industry among young talents. The role in the project is to test the digital twin software that will implement the AI solution. Software testing techniques is used to test the AI and the digital twin software systems.
Project duration: 2020-2022
Project Name: IMPAQCDT
IMPAQCDT – Improved Methods for Process and Quality Controls using Digital Tools
Project Description: IMPAQCDT aims to develop a novel approach based on a combination of fundamental knowledge with a concept of machine learning to accelerate the development of quality control methods and early assessment of the manufacturability of oligonucleotides. This activity will focus on formulating hybrid mathematical models based on combining the mechanistic description of chromatography and machine learning (ML) algorithms that produce digital copies of purification methods. The concepts of the model-based testing approach will be considered during this stage to ease the quality assurance process. This, in turn, will facilitate the test generation by generating the test cases from the model to verify the correctness. The testing strategy will consider the continuous evolution of the system and the unpredictable behavior of the ML algorithms.
Project duration: 2021-2024
Completed Projects
Project Name: CIT for Avocado
Project Description: The project aims to design and implement a new plugin to the open-source Avocado framework, which allows generating much more efficient test data. The new plugin will add the capability of Combinatorial Interaction Testing (CIT) to the framework by decreasing the number of test cases.
Official Page of Avocado Here
Github Repository Here
Project Name: Code-based Testing
Developing a test generation tool using an adaptive analysis process
Project Description: The goal of the project is to use code internal structures and analysis for test generation. The analyzed information is used in the test generation process in the developed tool. We introduced a new concept of Code Coverage-based Test Case Generation (CCTG) that changes the current practices by utilizing the code coverage analysis in the test generation process. Here, we follow gray-box testing by first generating an initial test suite to analyze the code coverage. We use the code coverage data to calculate the impact of each input parameter. Then we use this analysis information with a constraint solver to automate the generation of effective test suites. We applied this approach to a few real case studies. The project has been funded by the Red Hat lab and STILL group at the Czech Technical University in Prague.
Project duration: 2017-2019
Project Name: Smart TV app
Developing a framework for smart TV application Testing
Project Description: The goal of the project is to develop a framework to test the smart TV applications for IoT purposes. The aim of this project is to develop a fully automated Smart TV app testing framework based on a model-based testing approach. Throughout this project, we also aimed to create an open-source research community for smart TV app testing by providing the necessary guidelines, starting packages, tools, repositories, and documentation to motivate the collaboration in this direction. As an output of the project, we developed our EvoCreeper tool that can crawl the GUI of a smart TV app and then extract the necessary components to be included in a comprehensive model. The project has been funded by the Red Hat lab and STILL group at the Czech Technical University in Prague.
Project duration: 2017-2020
More details can be found in the project page here
Project Name: MCDCPSO
Design and Implementation of a Constraints Combinatorial Testing Strategy with Modified Condition/Decision Coverage (MC/DC) Using Fuzzy-based Adaptive Swarm Optimization
Project Description: This research addresses the integration of combinatorial set generation with MC/DC to fulfill the required interaction coverage as well as condition/decision coverage. The problem cannot be solved without using optimization and artificial intelligence techniques. To this end, this research proposes a new optimization technique using Fuzzy-based Adaptive Swarm Optimization. In doing so, the research attempt to design and develop a new strategy considering the aforementioned problem. The research has been funded by the Federal Department of Economic Affairs, Education and Research (EAER), Switzerland.
Project duration: 2015-2017
PSTG: Particle Swarm Test Generator
Project Description: Particle Swarm-based t-way Test Generator (PSTG) is designed for generating uniform and variable strength covering arrays. Unlike other existing AI-based t-way testing strategies, the lightweight computation of the particle swarm search process enables PSTG to support high interaction strengths of up to t=6. The performance of PSTG is evaluated using several sets of benchmark experiments. Comparatively, PSTG consistently outperforms its AI counterparts and other existing testing strategies as far as the size of the array is concerned. The project has been funded by the Ministry of higher education in Malaysia (MOHE).
Project duration: 2010-2012