BlackArc15 papers
Manufacturing & Supply Chain
Agents on the shopfloor and across the supply network: scheduling, inventory, sourcing, and disruption response.
- Papers
- 15
- Read time
- 4.0h
- Avg length
- 16m
A Large Language Model-based Multi-Agent Manufacturing System for Intelligent Shopfloors
Replace the hand-written heuristic rule that a factory's machines use to negotiate "who processes the next part" with an LLM that reads a plain-language bidding document and reasons its way to a choice — no simulator, no training, and it beats the heuristics it replaces.
- Read
- 18 min
- Source
- arXiv:2405.16887
Agent Manufacturing: Foundation-Model Agents as First-Class Industrial Entities
Every prior industrial revolution automated muscle or routine cognition but left the *coordinative* work of manufacturing — planning, scheduling, diagnosing, negotiating, and governing production — to humans; this paper argues foundation-model agents are the first technology aimed squarely at that layer, names the resulting paradigm "Agent Manufacturing," gives it a falsifiable definition, and is unusually blunt that the technology isn't there yet.
- Read
- 18 min
- Source
- arXiv:2605.24823
Agentic LLMs in the Supply Chain: Towards Autonomous Multi-Agent Consensus-Seeking
Replace a human phone call between a retailer and its supplier with two LLM agents that trade proposals back and forth inside a bounded range set by a real forecasting formula — and the resulting "negotiation" framework cuts both cost and the bullwhip effect more than either the formula alone or a bigger model would.
- Read
- 16 min
- Source
- arXiv:2411.10184
AI Agent Systems for Supply Chains: Structured Decision Prompts and Memory Retrieval
A well-written prompt (process description + a textbook safety-stock formula) lets a plain LLM multi-agent system hit the mathematically optimal inventory policy in the easy case, but it falls apart once demand changes over time — and the fix isn't more prompt tuning, it's giving each agent a per-tier memory of past (state, order, profit) triples it can retrieve by similarity, optionally pre-loaded from a cheap RL run.
- Read
- 16 min
- Source
- arXiv:2602.05524
AI Agents and Agentic AI – Navigating a Plethora of Concepts for Future Manufacturing
A field guide that untangles four overlapping buzzwords — AI agent, LLM-Agent, MLLM-Agent, and Agentic AI — into a clean evolutionary ladder, then maps each rung to a specific tier of manufacturing capability so you know which one you're actually selling or building.
- Read
- 13 min
- Source
- arXiv:2507.01376
Automating Supply Chain Disruption Monitoring via an Agentic AI Approach
Seven LLM agents turn a single news article into a verified, human-approved supplier-replacement plan by walking a Neo4j knowledge graph from Tier-4 all the way back to a company's direct suppliers — in under 4 minutes, for about 8 cents.
- Read
- 20 min
- Source
- arXiv:2601.09680
Deep Reinforcement Learning for Machine Scheduling: Methodology, the State-of-the-Art, and Future Directions
A field map of how deep reinforcement learning is being used to schedule factory machines — sorting every approach into four "brains" (plain neural nets, encoder-decoders, graph neural nets, and metaheuristic hybrids) and telling you which one wins on speed, scale, and generalization.
- Read
- 22 min
- Source
- arXiv:2310.03195
Demand Forecasting at Wayfair
Wayfair forecasts 18 months of monthly demand for 4 million products by blending a stable statistical "top-down" forecast with a responsive ML "bottom-up" forecast, mixing them per-item and per-horizon — and it earned tens of millions of dollars a year.
- Read
- 14 min
Fine-Tuning Vision-Language Model for Automated Engineering Drawing Information Extraction
A 0.23B-parameter open-source vision-language model, fully fine-tuned on just 400 expert-labeled engineering drawings, beats zero-shot GPT-4o and Claude-3.5-Sonnet by 30-52% on extracting tolerancing data — proving that for a narrow visual-extraction task, a tiny tuned model crushes a giant general one.
- Read
- 14 min
InvAgent: A Large Language Model based Multi-Agent System for Inventory Management in Supply Chains
Give one LLM to each stage of a supply chain, hand it a plain-English description of its inventory state and a "golden rule" strategy tip, and it manages orders competitively against trained reinforcement-learning policies — with zero training and a written explanation for every decision.
- Read
- 15 min
- Source
- arXiv:2407.11384
Large Language Models for Supply Chain Decisions
Put an LLM in front of your optimization solver — not in place of it — so non-technical planners can interrogate, stress-test, and re-run supply-chain models in plain English, cutting decision time from days to minutes while your proprietary data never leaves your walls.
- Read
- 16 min
- Source
- arXiv:2307.03875
Optimizing Multi-Tier Supply Chain Ordering with LNN+XGBoost: Mitigating the Bullwhip Effect
Pair a tiny, fast "liquid" neural net (great at reading time-series dynamics) with XGBoost (great at squeezing accuracy out of tabular features), then drive a profit-maximizing order simulator with the combined forecast — and you get more profit and less of the demand-amplification ("bullwhip") chaos that wrecks multi-tier supply chains.
- Read
- 16 min
- Source
- arXiv:2507.21383
Optimizing Supply Chain Networks with the Power of Graph Neural Networks
This paper takes a real-world supply-chain dataset (41 FMCG products, their dependencies, and 221 days of demand), defines "predict next demand for one product" as a benchmark task, and shows that graph-aware neural networks can forecast it far more accurately than a plain neural net — establishing baselines others can build on.
- Read
- 14 min
- Source
- arXiv:2501.06221
Real-Time Health Supply Chain Optimization Using Digital Twin Technology
Wire live IoT sensor data into a continuously-updating virtual model of a health supply chain so AI can forecast disruptions (spoilage, stockouts, equipment failure) and recommend fixes *before* they happen — instead of after.
- Read
- 14 min
Reliability and Effectiveness of Autonomous AI Agents in Supply Chain Management
They rebuilt the classic Beer Game with LLM agents instead of humans, found the agents can beat human teams on cost but are secretly unreliable — the same input can produce wildly different orders run to run, and that instability snowballs upstream — and showed that reinforcement-learning post-training (not prompting, not sampling more) is what actually fixes it.
- Read
- 17 min
- Source
- arXiv:2605.17036