FCRF: Flexible Constructivism Reflection for Long-Horizon Robotic Task Planning with Large Language Models

2025/07/20 by Yufan Song, Song, Yufan, Jiatao Zhang +11
Computer Science · Psychology · #AI-based Problem Solving and Planning #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Multimodal Machine Learning Applications #Robotics (cs.RO) #Social Robot Interaction and HRI

paper · pdf · doi:10.48550/arxiv.2507.14975

openalex publication_date 2025/07/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Autonomous error correction is critical for domestic robots to achieve reliable execution of complex long-horizon tasks. Prior work has explored self-reflection in Large Language Models (LLMs) for task planning error correction; however, existing methods are constrained by inflexible self-reflection mechanisms that limit their effectiveness. Motivated by these limitations and inspired by human cognitive adaptation, we propose the Flexible Constructivism Reflection Framework (FCRF), a novel Mentor-Actor architecture that enables LLMs to perform flexible self-reflection based on task difficulty, while constructively integrating historical valuable experience with failure lessons. We evaluated FCRF on diverse domestic tasks through simulation in AlfWorld and physical deployment in the real-world environment. Experimental results demonstrate that FCRF significantly improves overall performance and self-reflection flexibility in complex long-horizon robotic tasks.

Citations

• AutoManual: Constructing Instruction Manuals by LLM Agents via Interactive Environmental Learning
• Agent-Pro: Learning to Evolve via Policy-Level Reflection and Optimization
• Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models
• ExpeL: LLM Agents Are Experiential Learners
• Retroformer: Retrospective Large Language Agents with Policy Gradient Optimization
• AdaPlanner: Adaptive Planning from Feedback with Language Models
• Tree of Thoughts: Deliberate Problem Solving with Large Language Models
• Reflexion: Language Agents with Verbal Reinforcement Learning
• A Survey on In-context Learning
• Distributing Accountability, Not Capability: Phase Separation and the LLM Workflow Quadrant in Autonomous AI Agent Architectures
• Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
• Pre-Trained Language Models for Interactive Decision-Making
• ALFWorld: Aligning Text and Embodied Environments for Interactive Learning

Related

• Self-Corrective Task Planning by Inverse Prompting with Large Language Models
• Think, Act, Learn: A Framework for Autonomous Robotic Agents using Closed-Loop Large Language Models
• Enhancing Robustness in Language-Driven Robotics: A Modular Approach to Failure Reduction
• HiCRISP: An LLM-based Hierarchical Closed-Loop Robotic Intelligent Self-Correction Planner
• ReplanVLM: Replanning Robotic Tasks with Visual Language Models