Agentic AI and sustainable procurement: Rethinking anti-corrosion strategies in oil and gas
Year: 2025
Type: article
Abstract: Corrosion remains a persistent challenge in the oil and gas industry, causing infrastructure failures, escalating maintenance costs, and severe environmental risks. Recent innovations in anti-corrosive materials—such as nanotechnology-based coatings, corrosion-resistant alloys, and graphene composites—have been further accelerated by the rise of Artificial Intelligence (AI) in procurement. This paper introduces a next-generation framework for AI-driven procurement of anti-corrosive materials, integrating predictive analytics, generative AI, agentic AI, and blockchain-backed transparency. Generative AI enables dynamic supplier intelligence, automated sustainability reporting, and RFQ drafting, while agentic AI agents autonomously negotiate contracts, optimize sourcing cycles, and adapt procurement strategies in real time. Multimodal AI, when combined with IoT sensors and digital twins, facilitates predictive corrosion monitoring by analyzing sensor streams, visual inspection data, and supplier documents simultaneously. Furthermore, quantum AI simulations hold the potential to model corrosion resistance of emerging alloys under refinery-specific conditions, enabling more accurate material selection. Beyond cost savings and operational resilience, AI aligns procurement with sustainability goals by recommending low-VOC coatings, minimizing carbon footprints, and enabling circular economy practices. To ensure adoption, the paper emphasizes explainable AI (XAI) for procurement leaders, offering transparency into supplier rankings and material recommendations. By synthesizing advances in material science with cutting-edge AI, this study provides a strategic roadmap for resilient, sustainable, and autonomous procurement in the oil and gas sector.
Author: Waditwar, Prajkta
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7108221607
FWCI:
Cites:
Cited by:
Related to: