This disclosure describes monitoring and operating subsea well systems, such as to perform operations in the construction and control of targets in a subsea environment. A submerisble ROV that performs operations in the construction and control of targets (e.g., well completion components) in a subsea environment, the ROV has one or more imaging devices that capture data that is processed to provide information that assists in the control and operations of the ROV and/or well completion system while the ROV is subsea.

Systems and methods for mitigating cable twists for underwater cleaners are provided. Cable twist mitigation logic is stored in a memory associated with a pool or spa cleaner, and controls operation of the cleaner to mitigate cable twists. A sequence of cleaner orientations is retrieved from memory and compare to one or more pre-defined sequences known to contribute to cable twist. If the sequence of cleaner orientations matches the one or more pre-defined sequences, a twist angle accumulator is incremented by a pre-defined twist angle corresponding to the one or more pre-defined sequences. The system determines whether the cleaner is turning on a surface of a pool or spa, and if so, controls turning of the cleaner using an accumulated angle stored in the twist angle accumulator to mitigate cable twists. A user-definable bias value could also be applied by the system to further mitigate cable twists.

A method of training a machine learning, ML algorithm to control a watercraft is described. The watercraft is a submarine or a submersible submerged in water. The method is implemented, at least in part, by a computer, comprising a processor and a memory, aboard the watercraft. The method comprises: obtaining training data including respective sets of sensor signals, related to respective deterrents, and corresponding actions of a set of communicatively isolated watercraft, including a first watercraft; and training the ML algorithm comprising determining relationships between the respective sets of sensor signals and the corresponding actions of the watercraft of the set thereof. A method of controlling a watercraft by a trained ML algorithm is also described.

A method of training a machine learning, ML, algorithm to control a watercraft is described. The watercraft is a submarine or a submersible submerged in water. The method is implemented, at least in part, by a computer, comprising a processor and a memory, aboard the watercraft. The method comprises: obtaining training data including respective sets of environmental parameters and corresponding actions of a set of communicatively isolated watercraft, including a first watercraft; and training the ML algorithm comprising determining relationships between the respective sets of environmental parameters and the corresponding actions of the watercraft of the set thereof. A method of controlling a watercraft by a trained ML algorithm is also described.

The present disclosure provides a swimming pool robot and a control method of the swimming pool robot. The swimming pool robot includes: a vehicle body; a power switch and a water intake detection module both arranged at a bottom of the vehicle body; a first water outlet and a second water outlet arranged opposite to each other at a top of the vehicle body; and a battery, an MCU, a body state detection module for detecting a state of the vehicle body, an electric power reserve detection module for detecting an electric power reserve of the battery, a first driving mechanism for driving the first water outlet to discharge water, and a second driving mechanism for driving the second water outlet to discharge water are arranged inside of the vehicle body, wherein each of the battery, the power switch, the water intake detection module, the body state detection module, the electric power reserve detection module, the first drive mechanism and the second drive mechanism is individually connected to the MCU. According to the present disclosure, after the swimming pool robot is started up, the first driving mechanism or the second driving mechanism is opened only when the swimming pool robot in the water, and sinks into the bottom under the action of its own gravity and is in the static state, so as to drive the first water outlet or the second water outlet to discharge water, so that the swimming pool robot advances in the swimming pool to achieve the dirt suction function, the settling time of the swimming pool robot is reduced, thereby improving the cleaning efficiency.

A submersible vessel comprises a hull. The hull contains a plurality of subsystems. The subsystems comprise a sensor subsystem configured to sense potential target information regarding a potential target, a database subsystem configured to store target characterization information, a processing subsystem coupled to the sensing subsystem and to the database subsystem, and an ordnance subsystem. The processing subsystem is configured to process the potential target information according to the target characterization information to confirm the potential target as being a confirmed target. The ordnance subsystem comprises an ordnance magazine configured to store ordnance. The ordnance is deployable against the confirmed target, wherein the confirmed target is an aircraft.

Disclosed herein is a method of detecting stalled state of a dynamic pool equipment unit, comprising receiving a plurality of movement features relating to a dynamic pool equipment unit deployed in a water pool which are captured during a predefined sampling window and comprise (1) motion features of the pool equipment unit, and (2) operational features of electric motor(s) of the pool equipment unit, determining a movement pattern of the pool equipment unit using one or more statistical models applied to the plurality of movement features which are trained to estimate a stalled state of the pool equipment unit in which the pool equipment unit is pitched up and unable to advance on a slopped obstacle in the water pool, and causing the pool equipment unit to stop attempted advance in a current direction responsive to determining that the pool equipment unit is in the stalled state.

A wall collision U-turning method for a swimming pool cleaning robot, a swimming pool edge cleaning method, and an electronic device are provided. The wall collision U-turning method includes: controlling the swimming pool cleaning robot to move toward a wall of a swimming pool based on a preset orientation until the swimming pool cleaning robot collides with the wall of the swimming pool; controlling the swimming pool cleaning robot to move backward relative to the wall of the swimming pool until a spacing distance between the swimming pool cleaning robot and the wall of the swimming pool is sufficient for the swimming pool cleaning robot to perform a U-turn motion; and controlling the swimming pool cleaning robot to perform a U-turn relative to the wall of the swimming pool until the U-turn orientation of the swimming pool cleaning robot after completion of the U-turn is opposite to the preset orientation.

A swimming pool map boundary construction and swimming pool cleaning methods, an apparatus, and an electronic device are provided. A swimming pool cleaning robot is controlled to move forward and backward relative to each preset path in a swimming pool map that covers a swimming pool within a working area defined by the swimming pool, to determine two path endpoints of each preset path, and map boundaries of the swimming pool map are constructed based on the determined two path endpoints of each preset path in the swimming pool map, so that the construction of swimming pool map boundaries is more efficient, reasonable and accurate. Moreover, the swimming pool cleaning task performed based on the swimming pool map constructed by the above method can achieve comprehensive cleaning of the swimming pool and avoid omissions.

The present disclosure provides a method and apparatus for cleaning a swimming pool, and an electronic device and a storage medium thereof. The method includes: controlling a swimming pool cleaning robot to move, with respect to a grid map covering the swimming pool, in a work area defined in the swimming pool to establish a cleaning map including a plurality of cleaning blocks; and controlling the swimming pool cleaning robot to traverse each of the cleaning blocks in the cleaning map to clean the swimming pool. According to the present disclosure, the cleaning regions may be accurately constructed in the map according to the work area in the swimming pool, such that the swimming pool cleaning robot cleans the swimming pool in a quick, efficient, and energy-saving fashion.