Citation

Tan, Chi Jie and Hayashi, Eiji and Mowshowitz, Abbe and Lim, Way Soong (2026) Hybrid LLM-Genetic Programming: Supervising and Generating Diverse Behavior Trees for Autonomous Robot Evolution. Robotics, 15 (5). p. 98. ISSN 2218-6581

Text robotics-15-00098.pdf - Published Version Restricted to Repository staff only Download (7MB)

Abstract

first_pagesettingsOrder Article Reprints Open AccessArticle Hybrid LLM-Genetic Programming: Supervising and Generating Diverse Behavior Trees for Autonomous Robot Evolution by Chi Jie Tan 1,*ORCID,Eiji Hayashi 1,Abbe Mowshowitz 2ORCID andWay Soong Lim 3,* 1 Department of Creative Informatics, Kyushu Institute of Technology, Fukuoka 820-0067, Japan 2 Department of Computer Science, City College of New York, New York, NY 10031, USA 3 Faculty of Engineering and Technology, Multimedia University, Melaka 75450, Malaysia * Authors to whom correspondence should be addressed. Robotics 2026, 15(5), 98; https://doi.org/10.3390/robotics15050098 Submission received: 14 April 2026 / Revised: 2 May 2026 / Accepted: 9 May 2026 / Published: 11 May 2026 (This article belongs to the Special Issue AI-Powered Robotic Systems: Learning, Perception and Decision-Making) Downloadkeyboard_arrow_down Browse Figures Versions Notes Abstract Genetic Programming (GP) for evolving Behavior Trees (BTs) in autonomous robots often suffer from premature convergence, even when adaptive mutation mechanisms are employed. This paper proposes a novel hybrid framework that integrates Large Language Model (LLM) supervision into GP, in which the LLM performs holistic population analysis, adaptively regulates mutation rates, and generates targeted BTs to proactively address behavioral gaps in the evolving population. Unlike conventional evolutionary operators, the LLM introduces high-level semantic guidance by seeding underrepresented behavioral archetypes, thereby complementing stochastic genetic variation with structured exploration. The proposed method is evaluated in a Unity-based multi-task robotic simulation environment. Experimental results show that the hybrid approach significantly outperforms baseline GP with standard adaptive mutation, achieving a 71.7% faster emergence of Complete Robots, a 65.2% faster emergence of Excellent Robots, and a 28% increase in behavioral diversity. Notably, the two systems exhibit opposite mutation dynamics: the LLM-guided system progressively reduces mutation rates to promote exploitation, whereas the baseline maintains a high mutation rate. In addition, the LLM generates approximately 40 targeted BTs per run, proactively seeding the population with underrepresented behavioral archetypes. These performance gains are obtained with only a 13% computational overhead.

Item Type:	Article
Uncontrolled Keywords:	Genetic programming, behavior trees
Subjects:	T Technology > TJ Mechanical Engineering and Machinery > TJ212-225 Control engineering systems. Automatic machinery (General)
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	30 Jun 2026 08:36
Last Modified:	30 Jun 2026 08:36
URII:	http://shdl.mmu.edu.my/id/eprint/16162

Downloads

Edit (login required)