Managing Complex Failure Analysis Workflows with LLM-based Reasoning and Acting Agents

2025-06-18

AI Summary

To address the challenges of poor AI component interoperability and fragmented workflows in failure analysis (FA) laboratories, this paper proposes a Large Language Model–based Planning Agent (LPA) tailored for FA scenarios. The LPA integrates symbolic Hierarchical Task Network (HTN) planning, tool-augmented reasoning, and multi-source heterogeneous data interfaces to enable end-to-end autonomous orchestration of image-based anomaly detection, cross-source case retrieval, and annotated report generation. It introduces the first LLM-driven FA intelligence closed loop featuring explainability, traceability, and human-in-the-loop intervention capability. Evaluated in a real-world FA laboratory deployment, the system achieves a 42% increase in task completion rate, a 58% reduction in average analysis cycle time, and 91.3% accuracy on complex query responses. These results demonstrate substantial improvements in FA process automation and decision-making trustworthiness.

Technology Category

• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Multiagent Systems: Multiagent Planning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Failure Analysis (FA) is a highly intricate and knowledge-intensive process. The integration of AI components within the computational infrastructure of FA labs has the potential to automate a variety of tasks, including the detection of non-conformities in images, the retrieval of analogous cases from diverse data sources, and the generation of reports from annotated images. However, as the number of deployed AI models increases, the challenge lies in orchestrating these components into cohesive and efficient workflows that seamlessly integrate with the FA process. This paper investigates the design and implementation of a Large Language Model (LLM)-based Planning Agent (LPA) to assist FA engineers in solving their analysis cases. The LPA integrates LLMs with advanced planning capabilities and external tool utilization, enabling autonomous processing of complex queries, retrieval of relevant data from external systems, and generation of human-readable responses. Evaluation results demonstrate the agent's operational effectiveness and reliability in supporting FA tasks.

Problem

• Orchestrating AI models into cohesive Failure Analysis workflows
• Automating complex queries and data retrieval in Failure Analysis
• Enhancing reliability of AI-assisted Failure Analysis processes

Innovation

• LLM-based Planning Agent for workflow orchestration
• Autonomous complex query processing with LLMs
• Integration of external tools for data retrieval