<p>In Wireless Sensor Networks, clustered networks use cluster head (CH) nodes to send aggregated data to the sink, helping conserve energy. Ensuring fault tolerance is essential. A failure of the CH causes a complete communication breakdown. Rotating the CH role across sensor nodes is vital for maintaining load balance. Effectively managing CH is essential for network performance, as their failure can disrupt communication. In many cases, CH nodes are required to transfer data to distant sink nodes due to the inefficiency of clustering and CH selection algorithms. This study proposes a Grey Wolf Optimization approach for optimal CH selection and a relay-based adaptive dynamic data aggregation method to address these challenges. This method effectively adjusts to the network’s dynamics. In the proposed data aggregation method, sensors can send data directly to the sink if the hop count between them and the sink is lower than between the sensors and the CHs. Standard data aggregation at the CHs will be applied if the hop count exceeds this limit. This dynamic aggregation technique boosts network energy efficiency and data aggregation by adapting to various configurations. Experimental results show that the proposed technique outperforms standard clustering methods.</p>