HARNESS: Human-Agent Risk Navigation and Event Safety System for Proactive Hazard Forecasting in High-Risk DOE Environments

TL;DR

HARNESS addresses proactive hazard forecasting in high-risk DOE environments by coupling LLM-driven risk analysis with a modular, agentic RAG architecture. The system orchestrates data ingestion, summarization, retrieval, failure-mode analysis, hazard-control extraction, and policy alignment under a central Orchestrator Agent to produce auditable vulnerability reports. Key contributions include a smart retrieval pipeline, FEMA-based failure analysis, SBMS policy integration, and SME-in-the-loop refinement that enhances traceability and adaptability. Preliminary evaluation shows solid retrieval grounding and high-accuracy hazard reporting, with planned quantitative assessments of accuracy, SME agreement, and latency reduction to demonstrate practical impact.

Abstract

Operational safety at mission-critical work sites is a top priority given the complex and hazardous nature of daily tasks. This paper presents the Human-Agent Risk Navigation and Event Safety System (HARNESS), a modular AI framework designed to forecast hazardous events and analyze operational risks in U.S. Department of Energy (DOE) environments. HARNESS integrates Large Language Models (LLMs) with structured work data, historical event retrieval, and risk analysis to proactively identify potential hazards. A human-in-the-loop mechanism allows subject matter experts (SMEs) to refine predictions, creating an adaptive learning loop that enhances performance over time. By combining SME collaboration with iterative agentic reasoning, HARNESS improves the reliability and efficiency of predictive safety systems. Preliminary deployment shows promising results, with future work focusing on quantitative evaluation of accuracy, SME agreement, and decision latency reduction.

Figures (1)

• Figure 1: HARNESS system architecture.