System and method for fusion of sensor data to support autonomous maritime vessels
AssigneeLeidos, Inc.
InventorsNahid Sidki, Weicheng Shen, Suk Lee, Andrew Genduso
Life of this patent
Estimating the term…
Abstract
A system of modular components can be used with existing sensor suites to fuse data and determine the operating environment (surface contacts/tracks) for an autonomous marine vehicle and feed an autonomy decision engine to improve the vessel arbitration process in deciding which way to turn, how fast to go, obstacle avoidance, and mission monitoring. The system includes the ability to obey the set of navigation rules published by the International Maritime Organization, generally referred to as COLREGS (collision regulations).
Claims12
A process for identifying and tracking maritime vessel contacts by an autonomous maritime vehicle for use in autonomous navigation of the autonomous vehicle comprising:
- receiving multiple sensor data sets from multiple sensors associated with the autonomous maritime vehicle and fusing the sensor data by the data fusion architecture to determine a state of the current operating environment of the autonomous maritime vehicle, wherein a state of the current operating environment includes identification of one or more maritime vessel contacts and one or more tracks thereof within the operating environment by application of a Multi-Hypothesis Tracking (MHT) algorithm to the multiple sensor data sets;
- and
- processing results of the application of the MHT algorithm to the fused sensor data from the data fusion architecture including assessment of an operational state of one or more sensed maritime vessel contacts, computation of maritime vessel contact density within the current operating environment and generation of one or more collision alerts and using the processed results to navigate the autonomous maritime vehicle, wherein navigation includes directing the autonomous vehicle to perform one or more of the following actions turn, change speed, and avoid an obstacle.
An autonomous navigation system for directing an autonomous maritime vehicle, the system comprising:
- multiple sensors located on the autonomous maritime vehicle;
- a sensor data fusion architecture including a sensor processing module and a track processing module;
- wherein the sensor processing module receives data inputs from the multiple sensors, the data inputs providing information relevant to identification of maritime vessels within an operational range of the autonomous maritime vehicle and useful for determining a predicted track of each of the identified maritime vessels by a Multi-Hypothesis Tracking (MHT) algorithm of the track processing module;
- wherein the track processing module includes one or more tracking filters selected from the following group consisting of Kalman Filters, Interacting Multiple Model Filters and Particle Filters;
- and
- further wherein the track processing module is programmed to recognize when the information from the data inputs associated with the identified maritime vessels indicates multiple tracks for a single identified maritime vessel and combines the multiple tracks to a single predicted track for the single identified maritime vessel.
Dependencies are inferred from the claim text.