https://www.mdu.se/

Federated Learning (FL) has emerged as a novel paradigm within machine learning (ML) that allows multiple devices to collaboratively train a shared ML model without sharing their private data with a central server. FL has gained popularity across various applications by eliminating the necessity for centralized data storage, thereby improving the confidentiality of sensitive information. Among the new FL applications, this thesis focuses on Speech Emotion Recognition (SER), which involves the analysis of audio signals from human speech to identify patterns and classify the conveyed emotions. When SER is implemented within a FL framework, even though speech data remains on local devices, new privacy challenges emerge during the training phase and the exchange of SER model update parameters between servers and clients. These challenges encompass the potential for privacy leakage and adversarial attacks, including model inversion attacks and membership or property inference attacks, which can be conducted by unauthorized or malicious parties to exploit the shared SER model, compromising client data confidentiality and revealing sensitive information.

While several privacy-preserving solutions have been developed to mitigate potential breaches in FL architectures, those are too generic to be easily integrated into specific applications. Furthermore, incorporating existing privacy-preserving mechanisms into the FL framework can increase communication and computational overheads, which may, in turn, compromise data utility and learning performance.

This thesis aims to propose privacy-preserving methods in FL for emerging security-critical applications such as SER while addressing the challenges related to their effect on performance. First, we categorize and analyze recent research on privacy-preserving mechanisms in FL, with a focus on assessing their effects on FL performance and how to balance privacy and performance across various applications. Second, we design an optimized FL setup tailored to SER applications in order to evaluate effects on performance and overhead. Third, we design and develop privacy-preserving mechanisms within FL to safeguard against potential privacy threats while ensuring the confidentiality of clients' data. Finally, we propose and evaluate new methods for FL in SER and integrate them with appropriate privacy-preserving mechanisms to achieve an optimal balance of privacy with efficiency, accuracy, as well as communication and computation overhead.

In Industry 4.0, production, Information Technology (IT), and the Internet are combined. The nine technologies of Industry 4.0, Artificial Intelligence (AI) and Cyber-Physical System (CPS), are changing machines, strategies, processes, and maintenance.

In the first generation of maintenance, machines were run to failure, which is related to Corrective Maintenance. Systems for planning and control were implemented in the second generation, related to Predetermined Maintenance. Condition Based Maintenance (CBM) was presented in the third maintenance generation. Industry 4.0 places new demands on maintenance and different maintenance approaches are presented in previous research, such as Maintenance 4.0, Smart Maintenance and Self-Maintenance. This research focuses on smart maintenance technologies, using the nine technologies of Industry 4.0, such as Industrial Internet of Things (IIoT), and Big Data and Analytics, for machine connection, maintenance data collection, analysis of data, and making decisions using AI. CPS can be used to integrate the physical world, such as manufacturing machines, factory environment, material, people, and executions, with the cyber world, such as data analysis, apps, services, and decision-making.

Previous research presents several approaches to smart maintenance technologies. One problem is a lack of research regarding how smart maintenance technologies can be implemented to add benefits to the maintenance organization in line with company’s goal. Furthermore, previous research presents that further research is needed to support the manufacturing industry in what step an organization should take to implement smart maintenance technologies. In this research, four studies have been performed, which include literature reviews to obtain a clear overview of the research area of smart maintenance, as well as collected empirical data. The empirical data is collected from large companies and Small and Medium-sized Enterprises (SMEs), within the manufacturing industry, to obtain a clear overview of the manufacturing industry’ situation. The studies show that the manufacturing industry faces several challenges when implementing smart maintenance technologies, despite the concept of Industry 4.0 has been discussed for more than ten years. In this research, a conceptual implementation process is proposed, including challenges and enablers to consider when implementing smart maintenance technologies, as well as benefits of using smart maintenance technologies.