LLM-Based Behavior Tree Generation for Construction Machinery

2026-02-01

AI Summary

This work addresses the limited scalability of ROS2-TMS systems in complex construction environments, where manually designed behavior trees struggle to support coordinated automation across diverse types of construction machinery. To overcome this challenge, the authors propose a two-stage LLM-based approach for automatic behavior tree generation: first performing high-level task planning, then instantiating concrete behavior trees via structured templates enriched with parameters from the system database to ensure safety. A novel synchronization flag mechanism is introduced to enable safe multi-machine coordination. This study presents the first application of large language models to real-world multi-machine collaborative tasks in actual construction settings. Both simulation and field experiments demonstrate that the proposed method significantly enhances the scalability and practicality of automated coordination in dynamic construction scenarios.

Technology Category

• Multiagent Systems: Coordination and Collaboration
• Intelligent Robots: Behavior Learning & Control
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Earthwork operations are facing an increasing demand, while workforce aging and skill loss create a pressing need for automation. ROS2-TMS for Construction, a Cyber-Physical System framework designed to coordinate construction machinery, has been proposed for autonomous operation; however, its reliance on manually designed Behavior Trees (BTs) limits scalability, particularly in scenarios involving heterogeneous machine cooperation. Recent advances in large language models (LLMs) offer new opportunities for task planning and BT generation. However, most existing approaches remain confined to simulations or simple manipulators, with relatively few applications demonstrated in real-world contexts, such as complex construction sites involving multiple machines. This paper proposes an LLM-based workflow for BT generation, introducing synchronization flags to enable safe and cooperative operation. The workflow consists of two steps: high-level planning, where the LLM generates synchronization flags, and BT generation using structured templates. Safety is ensured by planning with parameters stored in the system database. The proposed method is validated in simulation and further demonstrated through real-world experiments, highlighting its potential to advance automation in civil engineering.

Problem

• Behavior Trees
• Construction Automation
• Large Language Models
• Heterogeneous Machine Cooperation
• Task Planning

Innovation

• LLM-based planning
• Behavior Tree generation
• synchronization flags
• construction automation
• heterogeneous machine cooperation