Digital Twins Below the Surface: Enhancing Underwater Teleoperation

TL;DR

The paper tackles underwater teleoperation challenges for ROVs by proposing an underwater digital twin that fuses real-time robot pose and environmental mapping with visualization and simulation. The system is built around three components — physical (BlueROV2 with ROS and ORB-SLAM3), digital (Unity-based interface with UWRobotics simulator), and bidirectional communication (ROS bridge) — enabling teleoperation, autonomous navigation, and testing. They validate the approach through real-data and simulation experiments, showing enhanced situational awareness, ability to replay operations, and autonomous navigation within the DT. Key findings include successful teleoperation data transmission, feasible path planning with RRT*, and a roadmap for deployment and improvement such as denser point clouds and real-world hardware experiments.

Abstract

Subsea exploration, inspection, and intervention operations heavily rely on remotely operated vehicles (ROVs). However, the inherent complexity of the underwater environment presents significant challenges to the operators of these vehicles. This paper delves into the challenges associated with navigation and maneuvering tasks in the teleoperation of ROVs, such as reduced situational awareness and heightened teleoperator workload. To address these challenges, we introduce an underwater Digital Twin (DT) system designed to enhance underwater teleoperation, enable autonomous navigation, support system monitoring, and facilitate system testing through simulation. Our approach involves a dynamic representation of the underwater robot and its environment using desktop virtual reality, as well as the integration of mapping, localization, path planning and simulation capabilities within the DT system. Our research demonstrates the system's adaptability, versatility and feasibility, highlighting significant challenges and, in turn, improving the teleoperators' situational awareness and reducing their workload.

Figures (6)

• Figure 1: DT interface featuring UUV visualization, point cloud seafloor recreation, and sensor data display.
• Figure 2: Digital Twin System Overview.
• Figure 3: Digital Twin interface visualization: a) first person viewpoint; b) third person viewpoint.
• Figure 4: Digital Twin teleoperation snapshots: a) Digital Twin interface; b) simulation environment.
• Figure 5: Octomap structure of mapped sea floor environment with the robot's planned path and waypoints (marked in green)