Balancing privacy and performance in the age of data-driven technologies
Haleh Hayati defended her PhD thesis at the Department of Mechanical Engineering on March 11th.
The research of introduces privacy-preserving mechanisms that prevent adversaries from extracting critical information from shared data without excessively compromising performance. These methods apply to remote monitoring, networked control systems, cloud computing, and artificial intelligence. By leveraging tools from information theory, control engineering, and differential privacy, the proposed solutions enable secure data sharing while preserving key functions like predictive analytics, machine learning, and industrial automation.
One of the most significant contributions of this research is an encoding mechanism that transforms data and algorithms before sharing them with cloud servers. While cloud computing provides powerful remote resources, transferring sensitive data to external servers poses security risks. This novel coding framework ensures privacy while preserving computational accuracy and efficiency.
Real-world applications
The effectiveness of these privacy-preserving techniques is demonstrated through real-world applications. In semiconductor manufacturing, AI-driven processes optimize production while protecting proprietary data. Similarly, in medical diagnostics, AI-powered systems improve patient care without exposing sensitive health information. These solutions ensure that industries can benefit from AI while maintaining data security.
Privacy-preserving technologies
With increasing reliance on AI and cloud computing, protecting sensitive information is more critical than ever. This research highlights the importance of privacy-preserving tools that empower industries to use advanced technologies securely. By addressing fundamental privacy concerns, Hayati鈥檚 thesis provides practical solutions for building secure and trustworthy digital infrastructure.
The findings contribute to the responsible development of privacy-preserving technologies that support both societal progress and individual rights. By demonstrating that strong data privacy and system efficiency can coexist, this research paves the way for AI and cloud computing to drive innovation without compromising security.
Title of PhD thesis: . Promotor: Prof. Nathan van de Wouw. Co-promotor: Dr. Carlos Murguia Rendon.