AAM-SEALS: Developing Aerial-Aquatic Manipulators in SEa, Air, and Land Simulator
William Yang, Karthikeya Kona, Yashveer Jain, Tomer Atzili, Abhinav Bhamidipati, Xiaomin Lin, Yantian Zha
TL;DR
The paper tackles the lack of tools for cross-medium cross-environment robotics by introducing Aerial-Aquatic Manipulators (AAMs) and the SEALS simulator, enabling integrated flight, swimming, and manipulation in sea, air, and land. It couples a dynamic AAM model with a control stack that handles changing CoG, and embeds SPH-PBD-based underwater dynamics alongside linear aerodynamics for realism. Key contributions include (i) a coupled AAM dynamics and CoG-aware control framework, (ii) SEALS, a photorealistic SEAL simulator with realistic air-water transitions, waves, underwater optics, and aquatic fauna, (iii) an aquatic-animal search-and-capture challenge and RLfD framework (SACfD) with demonstrations and visual RL, all released open-source to spur cross-domain robotics research. The work demonstrates realistic cross-medium simulations, precise control in SEALS, and RL-based learning capabilities, offering a practical platform for rapid TEVV and future physical AAM development, with future directions focusing on power modeling, additional manipulators, and scalable simulation.
Abstract
Current mobile manipulators and high-fidelity simulators lack the ability to seamlessly operate and simulate across integrated environments spanning sea, air, and land. To address this gap, we introduce Aerial-Aquatic Manipulators (AAMs) in SEa, Air, and Land Simulator (SEALS), a comprehensive and photorealistic simulator designed for AAMs to operate and learn in these diverse environments. The development of AAM-SEALS tackles several significant challenges, including the creation of integrated controllers for flying, swimming, and manipulation, and the high-fidelity simulation of aerial dynamics and hydrodynamics leveraging particle-based hydrodynamics. Our evaluation demonstrates smooth operation and photorealistic transitions across air, water, and their interfaces. We quantitatively validate the fidelity of particle-based hydrodynamics by comparing position-tracking errors across real-world and simulated systems. AAM-SEALS benefits a broad range of robotics communities, including robot learning, aerial robotics, underwater robotics, mobile manipulation, and robotic simulators. We will open-source our code and data to foster the advancement of research in these fields. The overview video is available at https://youtu.be/MbqIIrYvR78. Visit our project website at https://aam-seals.umd.edu for more details.