WasserMan: Benchmark for Underwater Manipulation Policy Learning
WasserMan: Benchmark for Underwater Manipulation Policy Learning
This paper introduces WasserMan, the first multi-task simulation benchmark dedicated to visuomotor policy learning for articulated underwater mechanisms.
The authors present WasserMan, a simulation benchmark for underwater manipulation policy learning that compares ACT, diffusion, and behavioral cloning methods while evaluating adaptation of a pretrained SmolVLA baseline under varied physical conditions.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S2
- code: Reported pending Open source S4
- dataset: not_reported
- paper: not_reported
- weights: not_reported
What remains unknown
Dataset licensing and commercial-use rights are not established here.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2610.04536v2 Announce Type: replace Abstract: Underwater contact manipulation couples visual decisions with vehicle stabilization and mechanism interaction. To our knowledge, WasserMan is the first multi-task simulation benchmark focused on visuomotor policy learning for manipulating articulated underwater mechanisms. The suite supports demonstration-based policy learning, alternative vehicle--arm platforms and
Open source S2
bimanual manipulation. We compare ACT, diffusion policies and behavioral cloning under matched sample budgets, and evaluate adaptation of a pretrained SmolVLA baseline. Fixed-policy studies vary water appearance, lighting, contact parameters, actuator delay, arm properties and mechanism pose. RGB feedback controls and matched observation-history and expert-trajectory ablations examine observation and training choice
Open source S3
s. Expert references establish task feasibility under the same physical success tests. The experiments reveal task-dependent visual transfer and pose sensitivity, while contact, tracking and actuator measurements distinguish execution responses that completion alone hides. Complete expert candidate streams make the initial states associated with demonstration selection inspectable. Versioned code, demonstrations, che
Open source S4
ckpoints and per-episode physical evidence support reproducible comparisons within the simulation.
Open source S5
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Confirm the required data type and collection setting with the buyer; this source does not establish a complete collection specification.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org