The integration of Internet of Things (IoT) technologies with autonomous vehicles presents a promising avenue for enhancing transportation safety, efficiency, and sustainability. This paper explores the application of federated learning (FL) to enable collaborative perception across fleets of self-driving vehicles using edge devices. We propose a framework that leverages IoT connectivity, federated learning, and edge computing to address challenges in data privacy and bandwidth constraints while harnessing the collective intelligence of vehicle fleets. The paper discusses fundamental principles and key components underpinning IoT-based collaborative perception systems, including connectivity, data collection, and automated decision-making. We identify critical applications such as improved object detection and tracking, high-definition environmental mapping, and traffic flow optimization. The potential benefits and challenges of implementing this technology are examined, along with future prospects for research and development. Our work contributes to the ongoing dialogue about the future of autonomous transportation and the role of collaborative technologies in shaping safer, more efficient, and sustainable urban mobility solutions.