Data Science Talk Series 

Title: Vehicle Computing - From Traditional Transportation to Computing on Wheels 

Abstract: With rapid advancements in edge computing, sensing, AI, and communication technologies, vehicles are undergoing a remarkable transformation. This talk will introduce a groundbreaking computing paradigm for the autonomous driving era—vehicle computing. This paradigm decouples data and control layers, creating an open computing platform that fosters multi-party collaboration and data sharing, breaking free from the constraints of traditional, closed vehicle systems. By doing so, vehicles transcend their conventional roles in transportation, evolving into versatile mobile computing platforms capable of supporting a wide range of advanced applications and third-party services. In addition, I will present a series of research efforts focused on developing reliable, scalable, and efficient edge-enabled connected vehicle applications. These include a vehicle-edge-cloud closed-loop framework that significantly reduces bandwidth usage while accelerating inference, a collaborative learning framework that enables both cooperative training and inference, in-transit noise mitigation, and the application of multimodal large language models (MLLMs) for autonomous driving.

Bio of the speaker: Sidi Lu is a tenure-track assistant professor of computer science at William & Mary. Her research interests include edge computing, vehicle computing, and applied AI & data science, with a focus on enhancing the reliability, scalability, security, and efficiency of connected and autonomous systems. She has collaborated closely with industry leaders such as Nokia Bell Labs, GM, Toyota, and Ford. She is the recipient of the NSF CRII Award, the Ralph H. Kummler Research Award, and the Michael E. Conrad Research Award. Her disk failure prediction software and related datasets have been downloaded by researchers from over 220 institutions worldwide since 2020.