FailureSensorIQ: A Multi-Choice QA Dataset for Understanding Sensor Relationships and Failure Modes

TL;DR

FailureSensorIQ introduces a multi-choice QA benchmark to evaluate large language models' ability to understand and reason about sensor data and failure modes in industrial settings, revealing strengths and weaknesses of current models.

Contribution

This work presents a novel MCQA benchmark, FailureSensorIQ, for assessing LLMs' reasoning on industrial sensor data and failure modes, along with analysis tools and a feature selection pipeline.

Findings

Closed-source models like GPT-4 approach expert-level performance.

Models show performance drops when faced with perturbations and distractions.

Significant knowledge gaps and fragility in current LLM reasoning capabilities.

Abstract

We introduce FailureSensorIQ, a novel Multi-Choice Question-Answering (MCQA) benchmarking system designed to assess the ability of Large Language Models (LLMs) to reason and understand complex, domain-specific scenarios in Industry 4.0. Unlike traditional QA benchmarks, our system focuses on multiple aspects of reasoning through failure modes, sensor data, and the relationships between them across various industrial assets. Through this work, we envision a paradigm shift where modeling decisions are not only data-driven using statistical tools like correlation analysis and significance tests, but also domain-driven by specialized LLMs which can reason about the key contributors and useful patterns that can be captured with feature engineering. We evaluate the Industrial knowledge of over a dozen LLMs-including GPT-4, Llama, and Mistral-on FailureSensorIQ from different lens using Perturbation-Uncertainty-Complexity analysis, Expert Evaluation study, Asset-Specific Knowledge Gap analysis, ReAct agent using external knowledge-bases. Even though closed-source models with strong reasoning capabilities approach expert-level performance, the comprehensive benchmark reveals a significant drop in performance that is fragile to perturbations, distractions, and inherent knowledge gaps in the models. We also provide a real-world case study of how LLMs can drive the modeling decisions on 3 different failure prediction datasets related to various assets. We release: (a) expert-curated MCQA for various industrial assets, (b) FailureSensorIQ benchmark and Hugging Face leaderboard based on MCQA built from non-textual data found in ISO documents, and (c) LLMFeatureSelector, an LLM-based feature selection scikit-learn pipeline. The software is available at https://github.com/IBM/FailureSensorIQ.