Edge computing is revolutionizing Internet of Things (IoT) networks by bringing data processing closer to the end devices, reducing reliance on centralized cloud systems. Traditional IoT architectures often experience problems with high latency, bandwidth, and security due to the massive amounts of data transmitted to remote (cloud) servers. By shifting computation to the edge infrastructure such as servers and gateways, computing enables faster decision-making, improves network efficiency, and enhances data security. The combination of Machine Learning (ML) with edge computing is further advancing IoT capabilities. Artificial Intelligence (AI)-powered edge infrastructure can analyze data right nearby end devices, which is highly effective in multiple scenarios, for example, for applications such as predictive maintenance in factories, intrusion detection in cybersecurity and traffic optimization in smart cities. Besides, processing data locally reduces network overloading and increases privacy by limiting unnecessary cloud transmissions. However, there are various challenges, which include managing limited resources of edge devices, ensuring good coordination between cloud and edge, and improving interoperability between IoT ecosystems. Ongoing research is focused on optimizing energy-efficient ML models, improving workload distribution, and strengthening security frameworks for edge computing. As more industries embrace edge-driven intelligence, this approach is set to play a crucial role in building faster, smarter, and more reliable IoT networks.