Examples relating to controlling extendable legs of a robotic device for use of a mechanical gripper are described herein. A robotic device may include one or more legs configured with a mechanical gripper coupled to the leg at a first position away from the distal end of the leg. The mechanical gripper may transition between the first position and a second position on the leg proximate to the distal end that enables the robotic device to use the mechanical gripper to perform grip operations. A control system of the robotic device may modify an orientation of the robotic device to at least position the robotic device on one or more legs other than the leg comprising the mechanical gripper, and transitions the mechanical gripper from the first position to the second position to perform the given grip operation of the input.

Examples relating to controlling extendable legs of a robotic device for use of a mechanical gripper are described herein. A robotic device may include one or more legs configured with a mechanical gripper coupled to the leg at a first position away from the distal end of the leg. The mechanical gripper may transition between the first position and a second position on the leg proximate to the distal end that enables the robotic device to use the mechanical gripper to perform grip operations. A control system of the robotic device may modify an orientation of the robotic device to at least position the robotic device on one or more legs other than the leg comprising the mechanical gripper, and transitions the mechanical gripper from the first position to the second position to perform the given grip operation of the input.

A system is disclosed including but not limited to a a jack up processor in data communication with each for dynamically balancing loads in real time on a plurality of legs supporting a jack up rig platform having a plurality of gear box motors on the plurality of legs. A processor reads data from sensors on gear box motors on the legs and selects a stored torque profile from a computer readable medium based on the load data from the sensors; and sends the torque profile to the plurality of gearboxes. A computer readable medium and neural network are disclosed for dynamically balancing loads on the plurality of legs in real time.

Provided are a robot accompaniment device and a four-legged robot (1) using the same. The robot accompaniment device includes a target module provided with an orienting assembly (3). The orienting assembly (3) is provided with a dual-antenna structure, and implements positioning and orienting between the accompanier and the accompanied, thus enabling a robot to be capable of recognizing the position, orientation and turning angle of an accompaniment object (2) relative to the robot, and moving and turning together with the accompaniment object (2).

Provided are a robot accompaniment device and a four-legged robot (1) using the same. The robot accompaniment device includes a target module provided with an orienting assembly (3). The orienting assembly (3) is provided with a dual-antenna structure, and implements positioning and orienting between the accompanier and the accompanied, thus enabling a robot to be capable of recognizing the position, orientation and turning angle of an accompaniment object (2) relative to the robot, and moving and turning together with the accompaniment object (2).

A bipedal vehicle 1 comprising two extendible output legs 5.sub.L, 5.sub.R (to support the vehicle), two foot holds 59.sub.L, 59.sub.R (to accept input movement from an operator) and a control system 253 comprising powered actuators to move the output legs in relation to the input movement to produce output movement.

A bipedal vehicle 1 comprising two extendible output legs 5.sub.L, 5.sub.R (to support the vehicle), two foot holds 59.sub.L, 59.sub.R (to accept input movement from an operator) and a control system 253 comprising powered actuators to move the output legs in relation to the input movement to produce output movement.

A method of mitigating slip conditions and estimating ground friction for a robot having a plurality of feet includes receiving a first coefficient of friction corresponding to a ground surface. The method also includes determining whether one of the plurality of feet is in contact with the ground surface, and when a first foot of the plurality feet is in contact with the ground surface, setting a second coefficient of friction associated with the first foot equal to the first coefficient of friction. The method also includes determining a measured velocity of the first foot relative to the ground surface, and adjusting the second coefficient of friction of the first foot based on the measured velocity of the foot. One of the plurality of feet of the robot applies a force on the ground surface based on the adjusted second coefficient of friction.

A method of robotic stepping includes determining a first step location error between a reference step location of a reference step path and a first potential step location of a first potential step path for a first leg of a robot, determining a first capture point error between a reference capture point location of the reference step path and a first potential capture point location of the first potential step path, determining a first score for the first potential step path based on the first step location error and the first capture point error, selecting the first potential step path based on comparing the first score for the first potential step path to a second score of a second potential step path, and instructing a movement of the first leg of the robot based on the first potential step path.

A method of robotic stepping includes determining a first step location error between a reference step location of a reference step path and a first potential step location of a first potential step path for a first leg of a robot, determining a first capture point error between a reference capture point location of the reference step path and a first potential capture point location of the first potential step path, determining a first score for the first potential step path based on the first step location error and the first capture point error, selecting the first potential step path based on comparing the first score for the first potential step path to a second score of a second potential step path, and instructing a movement of the first leg of the robot based on the first potential step path.