Because of the limits of single-path routing, the Multipath Routing (MR) approach is widely used in Mobile Ad-hoc Networks (MANET). In MANET, however, achieving energy-effective secure MR is a huge problem due to the need for trusted centralized experts and also limited resources. However, such systems have difficulties due to constraints such as packet loss rate. In this research, a low-power routing protocol that takes its cues from the well-known AOMDV protocol and a bio-inspired procedure, which is named as the Adapted Gear and Steering-based Rider Optimization Procedure (AGS-ROA) has been presented. Paths are found from the class of the fittest nodes with adequate energy for broadcast, lowering the chance of route failure and the rising sum of dead nodes brought on by larger data loads in the proposed AGS-ROA-AOMDV. After each transmission cycle, the AGS-ROA updating operator examines the nodes based on their remaining energy and changes the classes accordingly. Since gateways were chosen based on the total number of nodes, the EC was relatively high. So, a Fuzzy-based Clustering Algorithm (FCA) is suggested here. With this method, FCA is carried out using the Weighted Clustering Algorithm (WCA). Additionally, the FCA method estimates the total weight value using a number of characteristics (node degree, EC, latency, etc.) that are also used to assess the clustering quality.