On February 6, 2026, Ahmad Terra successfully defended his PhD thesis entitled “Explainable Artificial Intelligence for Telecommunications.”
Artificial Intelligence (AI) is a key driver of technological development in many industrial sectors. It is being embedded into many components of telecommunications networks to optimize their functionality in various ways. AI technologies are advancing rapidly, with increasingly sophisticated techniques being introduced. Therefore, understanding how an AI model operates and arrives at its output is crucial to ensure the integrity of the overall system. One way to achieve this is by applying Explainable Artificial Intelligence (XAI) techniques to generate information about the operation of an AI model.
This thesis develops and evaluates XAI techniques to improve the transparency of AI models. We took the opportunity to interview Ahmad and let him explain his research and what it contributes to.
Tell us what your research is about ?
- My research is about increasing the transparency of complex / blackbox AI models, specifically that are used in Telecommunications. It focuses on explainable AI where we want to understand why an AI comes to a particular decision/prediction. It also covers technical explanations on how the model works, as well as higher level explanations presented in natural language format.
How does your research contribute to solve a problem?
- The main contribution is to have a better understanding of AI operations. In a more specific context, we developed method to generate detailed explanations that can identify and mitigate bias. Our method can also be used to generate a meaningful (counterfactual) explanation. Furthermore, we also proposed method to evaluate textual explanations objectively.
Do you have you any particular memorable moment from your time as a PhD?
- It was very honorable to receive awards (Impact award and Key Contributor Award) from the company during my study. Additionally, I am glad to see that a method from my research was selected to be integrated in Ericsson product.
What challenges did you face during your PhD process?
- Technically, I have been exploring area that did not come to a fruition, I believe it is part of the process of completing my PhD. On the other hand, I live far from my home country and experienced a sequence of the hardest times in my life during the study.
What is ahead of you next?
- I am an industrial PhD in Ericsson, so I continue working as researcher, staying in Stockholm and may still have collaboration with KTH. So, I’m happy to be around.
Opponent:
Professor Kary Främling, Umeå University
Grading committee:
Associate Professor Sepideh Pashami, Halmstad University/RISE
Professor Kerstin Bach, Norwegian University of Science and Technology
Professor Fredrik Heintz, Linköping University
Spare committee member:
Associate Professor Hossein Azizpour
KTH Dissertation chair:
Associate Professor Martin Edin Grimheden, Head of Department Engineering Design
Supervisors:
Professor Martin Törngren, Engineering Design, KTH
Adjunct professor Rafia Inam, Ericsson and KTH
Link to “Explainable Artificial Intelligence for Telecommunications”