Federated edge-AI for reliable and privacy-preserving pipeline leak detection in drone swarms using neutrosophic sugeno-weber norms

Abstract The ability to monitor the safety of natural gas pipelines is guaranteed by leak detection. Systems are capable of responding quickly, and very precisely to events because delays during such events can lead to serious environmental consequences, hurt, damage, or even danger. Federated is a special framework that exists in this work. Leak Detection of Natural Gas pipelines Edge-A-enabled autonomous drone swarms will be real-time, where smart drones will be able to cooperate and reduce latency, keep sensitive data, and improve detection of anomalies in dynamic operational conditions such as complex decentralized control. system also needs advanced systems of decision-making that are capable of dealing with uncertainty, shifting goals, and information gaps or ambiguous information. The research will, in an attempt to achieve this, emphasize the Multi-Criteria Decision-Making (MCDM) methods that have been used over the years as an alternative method of analysis, which is systematic and founded on alternative performance measures. The precedent versions of MCDM, which applied the theory of fuzzy set, allowed the analysts to convey their judgments with vagueness and partial truth. As uncertainty and conflicting decision environments increased, however, neutrosophic sets were included to describe the degree of truth, falsity and indeterminacy on their own. This was a later representation that was refined to describe hesitation more by using ambiguous membership and non-membership functions of intuitionistic fuzzy sets (IFS). The combined paradigms led to Intuitionistic Neutrosophic Set (INS), a paradigm of powerful mathematics that can reflect the complexity and ambiguity of the decision-making problems of the real world. In this research, the INS framework is used with Sugeno-Weber (SW) aggregation operators to come up with a hybrid DM framework that is optimally designed to respond to the real-time leaks detection and assessment in pipeline networks. The proposed INS-SW solution is contrasted with the time-tested approach to the evaluation of performance, the Weighted Aggregated Sum Product Assessment (WASPAS), as it is easy to operate and can generate credible ranks. The comparative outcomes indicate that the INS-SW model can be better adapted to uncertain, interdependent and dynamic operation environments and is more robust and precise in that case. In general, the results suggest that the suggested framework adds to the fact and veracity of the drone-based leakage detection to a substantial degree and can provide a scalable and intelligent decision-making tool related to the imperative energy infrastructure. Besides this application in specific, the paper would also be applicable in developing uncertainty-sensitive decision science further, besides offering an insight into how to develop sustainable, intelligent, and resilient energy systems in future industrial processes.