
Voyis Launches Discovery Stereo Perception Series to Give Subsea Vehicles Real Time 3D Vision for Autonomy

Guest Contributor
Contributor
Voyis has launched the Discovery Stereo Perception Series, a new line of subsea stereo vision systems designed to improve pilot awareness and support autonomous underwater operations through real time 3D perception. The company frames the release as the next step after its Discovery Inspection Series, introduced three years ago, which has been used across hundreds of survey projects on a wide range of ROV platforms. Those deployments are presented as the operational base that informed the design priorities for autonomy-focused vision.
The Technology Gap Autonomy Has Not Solved Yet
Voyis argues that subsea autonomy has been limited by a mismatch between sensing modalities and the precision required for close-range work. Forward-looking sonar remains useful for longer-range detection and avoidance but does not provide the spatial resolution needed for fine manipulation or tight navigation near assets. Traditional single-camera setups deliver imagery but lack reliable depth perception, making it difficult to build repeatable autonomous behaviours that depend on understanding distance, geometry, and proximity in real time.
What the Perception Series Delivers On Vehicle
The new system replaces monocular camera configurations with stereo vision that produces true scale depth perception directly onboard the vehicle. Wide-angle stereo imaging paired with onboard processing generates dense depth maps and live point clouds intended to provide low-latency spatial awareness in complex underwater environments. The promise is that vehicles can shift from seeing a scene to understanding it in 3D in real time, enabling more reliable station keeping, closer asset approaches, and better support for manipulator tasks.
Read more: AI Speech Recognition Could Turn VHF Chatter into Clear Commands for Safer Tug and Pilot Operations
Operational Value for Pilots and Robotics Developers
Voyis positions the system as delivering three outputs at once, low-latency video for piloting, high-quality still images, and depth data that can be used for proximity monitoring and control frameworks. For operators, this is meant to improve positioning confidence and situational awareness during complex tasks. For developers, it is positioned as a foundation layer for autonomy, because depth perception can feed control loops and behaviour logic without relying only on sonar or post-processed visual reconstruction.
Additional Use in Rapid 3D Estimation Workflows
Beyond real-time autonomy support, Voyis says the Perception Series can also support photogrammetric workflows for general 3D estimation. Stereo imagery can be processed into indicative 3D models to help operators judge object size, geometry, and changes over time, providing a way to screen and prioritise targets when certified inspection-grade measurement is not required. The emphasis is on adding insight without adding workflow complexity or separate sensor packages.
Product Options and Platform Fit
The Discovery Stereo Perception Series is offered in two models, the P300 and P3000, intended to match different depth ratings and platform requirements. Both are described as providing wide field-of-view stereo imaging and onboard depth computation optimised for real-time performance, positioning the product as a compact, vehicle-integrated perception layer rather than a bulky add-on.
What This Signals for Subsea Autonomy
The launch reflects a broader move in subsea robotics toward higher autonomy and more precise close-range behaviours, where vehicles need dense spatial understanding rather than only imagery or long-range detection. If the system delivers consistent depth performance in real conditions, including variable visibility and challenging lighting, it could reduce a key barrier to scaling autonomous station keeping and manipulation tasks, and shift stereo vision from an inspection tool into a core navigation and control sensor for the next generation of underwater vehicles.

Guest Contributor
Contributor
This article was contributed by an external writer affiliated with our publication.




