DIGITAL.CSIC. Repositorio Institucional del CSIC
Datos de investigación
Set de datos (Dataset)
A Review and Case Study on Underwater Legged Robot Simulation [Dataset]
Please follow the links below to access the specific implementation files, launch scripts, and documentation for each simulation framework evaluated in the study:
1. Gazebo Sim Implementation: Contains the ROS 2 integration and standard rigid-body simulation setup for the ULR within the general-purpose Gazebo Sim environment, utilizing community-developed marine extensions.
2. Stonefish Implementation: Contains the deployment of the ULR within the marine-specific Stonefish simulator, demonstrating the framework's native underwater capabilities and advanced hydrodynamics.
3. NVIDIA Isaac Sim & "Isaac Sea" Plugin: Contains the AI-driven simulation setup and the source code for Isaac Sea, our custom GPU-accelerated hydrodynamics plugin developed to bridge the gap between high-fidelity marine physics and large-scale parallelization for Deep Reinforcement Learning (DRL).
4. Data Analysis & Benchmarking Results (Local Directory): This directory contains the raw quantitative data and Jupyter notebooks used to generate the figures and tables presented in the manuscript.
- Datasets (.csv): Includes aligned velocity metrics, drop test results, and computational resource measurements (RTF, GPU utilization, memory usage) for the terrestrial and hydrodynamic runs across all three simulators.
- Notebooks (.ipynb): Python scripts detailing the data alignment, scaling factors, and visualization logic used to produce the velocity comparison and drop test plots for the manuscript, This repository serves as the central data and code hub for the manuscript "A review and case study on underwater legged robotics simulation," submitted to Ocean Engineering.
To evaluate the transition from traditional Autonomous Underwater Vehicles (AUVs) to discrete, physical-contact Underwater Legged Robots (ULRs), this research benchmarks three distinct simulation architectures using a unified hexapod ULR case study (the SILVER2 platform).
Due to the fundamental architectural differences between the tested frameworks, the source code, implementation details, and benchmarking data have been divided into three dedicated repositories, This work was supported by:
The European Union’s H2020 program through the DIGI4ECO, SUN-BIO, REDRESS, MERLIN, and TERRATWIN projects [grant numbers GAP-101112883, GAP-101157493, GAP-101135492, GAP-01189796, GAP-101214557]
The Spanish Ministry of Science and Innovation (MICINN) through the AI4SEA project [grant number AIA2025-163346-C44]
The Spanish Ministry of Environment through the SMART-ME project [grant number PID2024-1553440B-C31]
The ``la Caixa'' Foundation [grant ID 100010434, fellowship code 161835], This work contributes to the Institut de Ciències del Mar "Severo Ochoa Centre of Excellence" accreditation CEX2024-001494-S funded by AEI 10.13039/501100011033 of the Spanish Ministry of Science and Innovation, Peer reviewed