Home // SIMUL 2025, The Seventeenth International Conference on Advances in System Modeling and Simulation // View article

Introduction of Reinforcement Learning into Automatic Stacking of Wave-dissipating Blocks

Authors:
Hao Min Chuah
Tatsuya Yamazaki

Keywords: wave-dissipating blocks; reinforcement learning; simulation; automatic stacking

Abstract:
Accurate and strategic placement of wave-dissipating blocks is essential for effective coastal protection structures. Current supervised learning-based approaches have achieved precise single-block placement. However, they inherently suffer from significant limitations, such as a lack of adaptability to environmental and structural changes, an inability to optimize sequences of multiple-blocks, and a heavy reliance on extensive pre-generated labeled data. This paper identifies the key limitations inherent in supervised Convolutional Neural Network methods and proposes a novel reinforcement learning (RL)-based approach to address these issues. By illustrating how RL naturally provides adaptability, strategic multi-block placement, and reduced reliance on labeled data, this early-stage idea is expected to contribute to the integration of simulation methodologies and machine learning approaches.

Pages: 40 to 42

Copyright: Copyright (c) IARIA, 2025

Publication date: September 28, 2025

Published in: conference

ISSN: 2308-4537

ISBN: 978-1-68558-300-2

Location: Lisbon, Portugal

Dates: from September 28, 2025 to October 2, 2025