3DGS-Driven Exocentric View System Significantly Boosts Subsea ROV Teleoperation Under High Latency
30 operators test a 3D Gaussian Splatting and vibrotactile suit for underwater ROV control.
Teleoperating remotely operated vehicles (ROVs) in flooded, cluttered infrastructure is fundamentally limited by narrow 2D egocentric views and subsea communication latency. To address this, a team led by Jing Du at the University of Florida developed a multimodal teleoperation architecture built on a ROS-Unity framework. The core innovation is the Dynamic Adaptive Viewpoint System (DAVS), which uses continuous optimization and real-time 3D Gaussian Splatting (3DGS) to synthesize an occlusion-free exocentric viewpoint from onboard state estimation, decoupling proactive spatial planning from reactive boundary avoidance. Additionally, a torso-mounted vibrotactile suit provides intuitive haptic proximity cues based on local obstacle clearance, reducing sensory workload.
In a controlled human-subject study (N=30) using a BlueROV2 navigating a complex simulated underwater facility, the team compared three interaction modalities (Egocentric, Haptic, Exocentric) under four communication delays (0.0-1.0 s). Performance was measured using behavioral metrics and functional near-infrared spectroscopy (fNIRS) to assess prefrontal activation. Results show that reactive haptic feedback improves path adherence under minimal delay, whereas the 3DGS-driven exocentric visualization provides superior resilience under severe latency (0.5-1.0 s), significantly outperforming the other modalities. fNIRS revealed a cognitive disengagement effect: increasing latency during conventional egocentric teleoperation overloaded working memory and reduced prefrontal activation, while the proactive spatial context of DAVS sustained executive control.
- Uses real-time 3D Gaussian Splatting (3DGS) to synthesize an occlusion-free exocentric viewpoint for subsea ROV teleoperation.
- Adds a torso-mounted vibrotactile suit that maps obstacle proximity to intuitive haptic cues, reducing cognitive load.
- fNIRS data shows DAVS maintains prefrontal executive control under 0.5-1.0s latency, while egocentric views cause cognitive disengagement.
Why It Matters
Spatially grounded multimodal assistance can dramatically improve operator performance and cognitive endurance in latency-degraded underwater teleoperation.