With edge computing on the ROV, the architecture is not burdened with providing for high-bandwidth communications between the operator and robot. Instead of raw video, sonar, sensor data, and command and control signals moving up and down a tether, situational awareness and work instructions can move instead. This allows us to consider using low-bandwidth and high-latency links, like acoustic and satellite links instead of fiber optic and Ethernet tethers.
Of course, with edge processing our ROVs come alive. Data processing, navigation, and control on the vehicle allow for decision making on the edge as well. This paves the way for redundancy, failure mitigation, and the platform required to integrate new levels of autonomy. With a modular and vehicle-agnostic autonomy solution, a “brain in a box”, addressing future requirements for residency, uncrewed operations, and extended reach is within reach of any manufacturer, service provider, or system integrator.
OPENSEA Edge combines the power of OPENSEA with a edge computing package to provide an easy solution for system integrators or vehicle developers moving towards full subsea and untethered autonomy. OPENSEA Edge presents the tried-and-true navigation, vehicle control, and autonomy solutions of OPENSEA with new OPENSEA technologies for objective-based autonomy, perception system integration, and low-bandwidth communications as a platform solution that can port to any vehicle system and support a wide range of applications. The OPENSEA Edge system is based on two NVIDIA edge processors and a Gigabit Ethernet network architecture while providing subsea connector interfaces for the robot, video sensors, sonar sensors, navigation sensors, and acoustic communications.
Edge computing places computational processes, the processing and consumption of data, as near to the source of the data as possible. In the context of ocean robots, it places the software responsible for processing and acting on sensor data on the robot itself versus with the operator. For many applications this is nothing new, but as Remotely Operated Vehicles (ROVs) find relevance in a world requiring persistence, residency, and long-range standoff deployments, ROV architecture has to abandon the “topside” architectures in favor of edge processing. ROVs have to develop brains.
- Platform agnostic
- Provides open architecture platform OPENSEA
- Provides over-the-horizon command and control
- Provides advance autonomy capabilities
- Room for other commercial off-the-shelf AI and ML engines