Amidst the swift advancement of technologies encompassing sensors, wireless communication, and cloud computing, the Internet of Things (IoT) connects various physical devices, sensors, and systems through the internet, enabling smart interconnection and data sharing. Wireless Sensor Networks (WSN) are a key component of the IoT, responsible for data collection and preliminary processing. WSN consists of numerous autonomous sensor nodes that transmit data to a central system or the cloud via wireless communication, commonly used for monitoring environmental information such as temperature, humidity, and vibration. However, since sensor nodes are typically powered by limited batteries, the network's lifespan and energy consumption balance become major challenges in research. Therefore, developing efficient clustering algorithms to extend the network's lifespan is crucial for the advancement of WSN. The LEACH (Low-Energy Adaptive Clustering Hierarchy) protocol is a classic adaptive clustering routing protocol designed for Wireless Sensor Networks (WSN), aimed at extending network lifetime by reducing energy consumption. It achieves load balancing within the network by dynamically selecting Cluster Head (CH) nodes, thereby decreasing the energy consumption of individual nodes. However, LEACH encounters issues such as data loss and network failures due to Cluster Head malfunctions and energy depletion. In response to these challenges, this paper proposes RLEACH, which particularly addresses the problem of Cluster Head energy depletion leading to network failures. Instead of adhering to a fixed number of data transmission rounds, RLEACH flexibly modifies the transmission rounds based on the energy levels of the Cluster Heads. Data transmission occurs only when the energy of the Cluster Head meets the requirements for all member nodes to transmit data in the next phase; otherwise, the network will initiate a re-clustering process. A comparison between the proposed RLEACH and LEACH demonstrates improvements of 119%, 148%, and 141% in terms of First Node Death (FND), Half Node Death (HND), and Last Node Death (LND), respectively. Additionally, RLEACH shows significant enhancements in energy consumption per round and overall network energy residual.