Introducing Marine Tensor™
Developed exclusively by Marine Thinking, Marine Tensor™ utilizes AI recognition, machine learning, and in-depth analysis of multiple sensor inputs such as optical, LiDAR, radar, and sonar to create full situational awareness of its environment in real-time. Its AI algorithm processes the collective data to calculate the optimal operational strategy to complete the mission task assigned by the control system.
Autonomous Navigation
Autopilot Navigation
Set your target locations and Marine Tensor™ will calculate and implement the optimal navigation strategy.
Obstacle Detection & Avoidance
Automatically detect obstacles, adjust course, and return-to-route in real-time without any risk to the mission.
Object Recognition & Tracking
Identify objects of interest and track their position in real-time.
Multi-tiered communication
Optimal Network Selection
Wi-fi, cellular network, satellite, or even through a tethered connection – automatically apply the optimal network connection based on the current environmental situation.
Real Time Data Streaming
Watch each mission unfold in real-time and make informed decisions with live data.
Remote Mission Control
Remotely assign new mission tasks, adjust instructions mid-mission, or take full control of the vessel on the fly.
SWARM Operations
Create a network of interconnected USVs – communicating and operating in sync with each other to complete the assigned mission task.
AI & Machine Learning Capability
To process and understand multiple layers of data in real-time, Marine Tensor™ utilizes Edge AI. This processing capability combines edge computing with artificial intelligence, which allows Marine Tensor™ to run its AI and machine learning algorithms directly from the surface vessel (local device). By hosting and processing data at the point of collection, we gain the ability to perform independent decision-making in real-time without depending on the cloud.
Using high-quality labeled datasets to train Marine Tensor™, its machine learning models learn to identify data patterns and make calculated decisions. When training the initial models, we established a high learning base at the outset by factoring in edge cases and varying operating conditions.
With each task performed, Marine Tensor™ continuously strengthens its predictive capabilities and confidence as it extracts more data patterns along the way.
Before any changes are applied to the AI and machine learning algorithms, all proposed modifications are run through a virtual simulation managed by Marine Tensor™ and the control system. Replicating the real-world environment, vigorous test simulations are applied to its digital twin. This allows us to train and improve the algorithms at a much faster pace and across infinite operational scenarios.
Interoperability
As an interoperability solution, Marine Tensor™ can serve as a central connection hub to uncrewed crafts above (UAV), below (ROV & AUV), and at surface level (USV). Utilizing its AI algorithms to automatically process and share data in real-time, the Marine Tensor™ architecture is designed to support multiple data sources simultaneously.
Sensor Fusion
Construct a highly detailed maps of the environment and identify objects of interest by combining data sets from a wide range of sensor technologies, including sonar, radar, LiDAR, optical, and thermal.
With full situational awareness of the vessel’s position and direction, Marine Tensor™ can continuously calculate and implement the optimal navigation strategy to complete its task.
remote payload Release
Whether it is an emergency situation or a routine task, remotely deliver important payloads to an assigned target location with pinpoint accuracy.
