Research

Paper

AI LLM March 18, 2026

PowerDAG: Reliable Agentic AI System for Automating Distribution Grid Analysis

Authors

Emmanuel O. Badmus, Amritanshu Pandey

Abstract

This paper introduces PowerDAG, an agentic AI system for automating complex distribution-grid analysis. We address the reliability challenges of state-of-the-art agentic systems in automating complex engineering workflows by introducing two innovative active mechanisms: (i) \textbf{adaptive retrieval}, which uses a similarity-decay cutoff algorithm to dynamically select the most relevant annotated exemplars as context, and (ii) \textbf{just-in-time (JIT) supervision}, which actively intercepts and corrects tool-usage violations during execution. On a benchmark of unseen distribution grid analysis queries, PowerDAG achieves a 100\% success rate with GPT-5.2 and 94.4--96.7\% with smaller open-source models, outperforming base ReAct (41--88\%), LangChain (30--90\%), and CrewAI (9--41\%) baselines by margins of 6--50 percentage points.

Metadata

arXiv ID: 2603.17418
Provider: ARXIV
Primary Category: eess.SY
Published: 2026-03-18
Fetched: 2026-03-19 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.17418v1</id>\n    <title>PowerDAG: Reliable Agentic AI System for Automating Distribution Grid Analysis</title>\n    <updated>2026-03-18T06:52:47Z</updated>\n    <link href='https://arxiv.org/abs/2603.17418v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.17418v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>This paper introduces PowerDAG, an agentic AI system for automating complex distribution-grid analysis. We address the reliability challenges of state-of-the-art agentic systems in automating complex engineering workflows by introducing two innovative active mechanisms: (i) \\textbf{adaptive retrieval}, which uses a similarity-decay cutoff algorithm to dynamically select the most relevant annotated exemplars as context, and (ii) \\textbf{just-in-time (JIT) supervision}, which actively intercepts and corrects tool-usage violations during execution. On a benchmark of unseen distribution grid analysis queries, PowerDAG achieves a 100\\% success rate with GPT-5.2 and 94.4--96.7\\% with smaller open-source models, outperforming base ReAct (41--88\\%), LangChain (30--90\\%), and CrewAI (9--41\\%) baselines by margins of 6--50 percentage points.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='eess.SY'/>\n    <published>2026-03-18T06:52:47Z</published>\n    <arxiv:primary_category term='eess.SY'/>\n    <author>\n      <name>Emmanuel O. Badmus</name>\n    </author>\n    <author>\n      <name>Amritanshu Pandey</name>\n    </author>\n  </entry>"
}