About

• Join our Team About this opportunity: Mobile network maintenance costs are a major part of operating costs of Communication Service Providers (CSPs). Site visits can impact maintenance costs negatively if visits are unplanned or frequent. Accordingly, ability to predict hardware failures before they happen will allow CSPs to move from unplanned site visits to planned ones where they can aggregate actions to minimize site visits. Predictive maintenance will use historical data to predict when equipment is likely to fail, allowing CSPs to schedule site visits more effectively to prevent hardware failure and minimize downtime What you will do: This thesis is divided into several steps with the end goal of showing a prototype of how hardware failures can be predicted. The following steps are envisioned as part of the thesis work: Decide on hardware failures in scope and data that will be used in the analysis to derive patterns. Leveraging collected historical data, use machine learning algorithms to predict when and why a specific HW is likely to fail. Analyze results of predicted failures and evaluate performance of different machine learning models. Optionally, design AI agent to automate the complete workflow and summarize results. The thesis will be concluded with a result presentation for the Ericsson research team. The skills you bring: This project aims at students in data science, computer science, computer engineering or similar. 1 Student, 30 hp Ericsson AB, Kista Preferred start: Jan 2026