Giovanni Gogliettino
TUTORIAL
AI on the Edge as a way to reduce power consumption and improve sustainability: two embedded application case studies
Giovanni Gogliettino
Technical Staff at STMicroelectronics
ABSTRACT
AI power consumption is increasing due to both the growing performance of machine learning models and the rising demand from users. One way to reduce this consumption is to move some processing to the edge, i.e., to embedded devices. In this talk, we will discuss the application of machine learning algorithms in embedded systems, focusing on their use in Global Navigation Satellite Systems (GNSS) receivers. Two studies will be described and analyzed: the first presents a machine learning framework aimed at improving GNSS functional safety by detecting and mitigating system anomalies, thereby enhancing the reliability of satellite-based navigation. Then, we will explore recent 2024 research leveraging reinforcement learning to optimize GNSS satellite signal acquisition, demonstrating how adaptive search strategies can improve performance in challenging environments. Together, these studies illustrate the potential of artificial intelligence in advancing GNSS robustness and efficiency, with implications for automotive and other critical applications.
SPEAKER BIOGRAPHY
Giovanni Gogliettino is a Senior Member of Technical Staff at STMicroelectronics, where he is involved in software research and development in the GNSS field. He obtained his master's degree in Computer Engineering from the University of Naples "Federico II" in 1999. In 2000, he began working at the IPM group SpA , where he dealt with cryptography and digital signatures. From 2002 to 2008, he worked for Incard SpA, where he initially focused on cryptography and digital signatures and later on the development of secure operating systems for smart cards. Since 2008, he has been working at STMicroelectronics.
