Large Language Models in the Loop: A Stability- and Network-Aware Survey in Networked Control, Cyber-Physical, and Multi-Agent Systems
eess.SY, cs.AI, cs.SY
Submitted: 2026-09-15
Updated: 2026-09-15
License: http://creativecommons.org/licenses/by/4.0/
The gist: Modern networked control systems (NCSs), cyber-physical systems (CPSs), and complex multi-agent network systems (CNSs) increasingly rely on large language models (LLMs) for high-level decision-making.
Terminology
Abstract
Modern networked control systems (NCSs), cyber-physical systems (CPSs), and complex multi-agent network systems (CNSs) increasingly rely on large language models (LLMs) for high-level decision-making. However, the slow, stochastic nature of LLMs directly conflicts with the strict stability and safety guarantees required by these physical systems. This survey presents a unified analysis of how LLMs can be admitted into the control loop of NCS, CPS, and CNS without compromising closed-loop guarantees. We organize this around a core principle: the LLM operates as a slow supervisor adjusting high-level goals and constraints, while a fast, certified inner loop maintains physical stability. Under this framework, LLM integration maps directly to classical networked control challenges, where inference latency acts as delay, API failures as packet dropouts, tokenization as quantization, and hallucinations as bounded disturbances. We assess current developments across all these three domains, highlighting that rising model capabilities are frequently accompanied by a drop in formal safety assurances. Finally, we propose concrete future research directions, identifying the widespread lack of formal stability proofs as the field's central open problem.
Sources
- Language Models are Few-Shot Learners
- OpenVLA: An Open-Source Vision-Language-Action Model
- Large Language Models for Power System Applications: A Comprehensive Literature Survey
- On the Opportunities and Risks of Foundation Models
- Capabilities of Large Language Models in Control Engineering: A Benchmark Study on GPT-4, Claude 3 Opus, and Gemini 1.0 Ultra
- MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework
- AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
- LLM+P: Empowering Large Language Models with Optimal Planning Proficiency
- ReAct: Synergizing Reasoning and Acting in Language Models
- Translating Natural Language to Planning Goals with Large-Language Models
- Neuro Symbolic Reasoning for Planning: Counterexample Guided Inductive Synthesis using Large Language Models and Satisfiability Solving
- ControlAgent: Automating Control System Design via Novel Integration of LLM Agents and Domain Expertise
- Eureka: Human-Level Reward Design via Coding Large Language Models
- Text2Reward: Reward Shaping with Language Models for Reinforcement Learning
- Language to Rewards for Robotic Skill Synthesis
- Spec2Control: Automating PLC/DCS Control-Logic Engineering from Natural Language Requirements with LLMs - A Multi-Plant Evaluation
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- Pre-Trained Large Language Models for Industrial Control
- Configuration Validation with Large Language Models
- LogLLM: Log-based Anomaly Detection Using Large Language Models
Related papers
- One Request, Multiple Experts: LLM Orchestrates Domain Specific Models via Adaptive Task Routing
- A Geometric Decision Procedure for STL Feasibility and Repair
- Submodular Multi-Agent Policy Learning for Online Distributed Task Allocation in Open Multi-Agent Systems
- Policy-Level Recursive Self-Improvement for Embodied AI with a Criticality World Model
- Minimal Experiments for Robust Stabilization: Information, Spectral Geometry, and Duration
- Decentralized Power-Optimal Coordination for Spacecraft Swarms Using Time-Varying Magnetorquer Actuation