PARAM: Prescriptive Agents Based on RAG for Automated Maintenance
Paper: Prescriptive Agents based on RAG for Automated Maintenance (PARAM) Authors: Chitranshu Harbola, Anupam Purwar Published: July 2025, arXiv Link: https://arxiv.org/abs/2508.04714 Category: Maintenance, Repair, Manufacturing
Key points
- PARAM applies retrieval-augmented generation (RAG) together with LLM-based agents to industrial maintenance, aiming to produce prescriptive (actionable) maintenance guidance rather than merely descriptive outputs.
- The work sits at the intersection of agentic search and large-scale search/recommendation systems: instead of a fixed retrieve-then-generate pipeline, retrieval frequency, strategy, and tool calls become part of the agent's decision space.
- The entry frames RAG/agentic search as evolving external knowledge access from "one-shot retrieval" into an iterative, verifiable, plannable process.
- Datasets/benchmarks commonly used in this space: MS MARCO, BEIR, Natural Questions, and domain-specific industrial corpora.
- Metrics: nDCG@10, MRR, Recall@k, task success rate, latency, and token cost.
- Baselines: BM25, dense retrieval, cross-encoder reranking, retrieval-free LLMs, commercial search APIs.
- Note: specific numerical results for PARAM should be checked against the original PDF; this entry summarizes design logic from the abstract and public metadata.
- A Compliance-Preserving Retrieval System for Aircraft MRO Task Search (arXiv:2511.15383)
- Bridging Industrial Expertise and XR with LLM-Powered Conversational Agents (arXiv:2504.05527)
- Enhancing Manufacturing Knowledge Access with LLMs and Context-aware Prompting (arXiv:2507.22619)
- MetalMind: knowledge-graph-driven human-centric knowledge system
- Optimizing Aerospace Product Maintenance with Multi-Modal Knowledge
- IR — Information Retrieval
- RAG — Retrieval-Augmented Generation
- nDCG — Normalized Discounted Cumulative Gain, a ranking-quality metric
- Agentic Search — modeling search as sequential decision-making and tool calling by an agent
- Gen-IR — Generative Information Retrieval
- Original paper: https://arxiv.org/abs/2508.04714
Context within agentic search
Neural information retrieval has progressed from BM25 to dense dual-tower retrievers, cross-encoders, late interaction, generative retrieval, and finally LLM agents. Each generation trades off efficiency, effectiveness, and maintainability. In the LLM era, new variables include inference budget and action space (whether to retrieve, how many times, and which tools to call). Evaluation correspondingly shifts from static metrics like nDCG@10 and MRR toward task success rate, citation accuracy, and multi-hop reasoning-chain completeness.
Typical method pipeline (as described in this class of systems)
1. Input & representation — encode queries, documents, and user/machine context into dense or sparse representations or structured prompts. 2. Core modules — retriever, reranker, planner, memory, and tool interfaces, composed sequentially or in parallel. 3. Learning strategy — supervised fine-tuning, contrastive learning, distillation, reinforcement learning, synthetic data bootstrapping. 4. Inference strategy — single-pass or iterative retrieval, parallel sub-queries, early stopping, and budget control.
Evaluation considerations
Engineering checklist
| Item | Question | Recommendation | |------|----------|----------------| | Data | PII in training/index? Version control? | Partitioned indexes, anonymization, rollback-capable embedding versions | | Latency | p99 budget? Retrieval steps? | Cascades + early stopping, hot-query caching, async reranking | | Quality | Do offline gains translate online? | Interleaving experiments, human audits, citation checks | | Safety | Poisoning/bias from open retrieval? | Source whitelisting, adversarial detection, output filtering | | Cost | Tokens/GPU per query? | Small-model routing, distillation, hybrid sparse+dense retrieval |