The Internet of Things (IoT) has many compelling applications in our daily lives. With the explosion of IoT devices and various applications, the demands on the performance, reliability, and security of IoT networks are higher than ever. Current end-host-based or centralized control frameworks generate excessive computational and communication overhead, and the dynamic response of IoT networks is sluggish and clumsy. Recently, with the advancement of programmable network hardware, it has become possible to implement IoT network functions inside the IoT network. However, current in-network schemes largely rely on manual processes, which exhibit poor robustness, flexibility, and scalability. Therefore, in this chapter, we present a new IoT network intelligent control architecture, in-network intelligence control. We design intelligent in-network devices that can automatically adapt to IoT network dynamics by leveraging powerful machine learning adaptive abilities. In addition, to enhance the collaboration among distributed in-network devices, a centralized management plane is introduced to ease the training process of distributed switches. To demonstrate the technical feasibility and performance advantage of our architecture, we present three use cases: in-network load balance, in-network congestion control, and in-network DDoS detection.