Context-aware LLM-based Safe Control Against Latent Risks

TL;DR

This paper introduces a comprehensive framework combining LLMs, optimization, and control to enable safe, context-aware autonomous decision-making that effectively manages latent risks in complex tasks.

Contribution

It presents an integrated multi-layer approach that decomposes tasks, refines subtasks through reasoning and optimization, and iteratively improves safety and performance in autonomous systems.

Findings

Successfully applied to robot arm and vehicle simulations

Demonstrates effective risk-aware decision-making

Improves learning efficiency of complex behaviors

Abstract

Autonomous control systems face significant challenges in performing complex tasks in the presence of latent risks. To address this, we propose an integrated framework that combines Large Language Models (LLMs), numerical optimization, and optimization-based control to facilitate efficient subtask learning while ensuring safety against latent risks. The framework decomposes complex tasks into a sequence of context-aware subtasks that account for latent risks. These subtasks and their parameters are then refined through a multi-time-scale process: high-layer multi-turn in-context learning, mid-layer LLM Chain-of-Thought reasoning and numerical optimization, and low-layer model predictive control. The framework iteratively improves decisions by leveraging qualitative feedback and optimized trajectory data from lower-layer optimization processes and a physics simulator. We validate the proposed framework through simulated case studies involving robot arm and autonomous vehicle scenarios. The experiments demonstrate that the proposed framework can mediate actions based on the context and latent risks and learn complex behaviors efficiently.