Boston Dynamics, Inc.

Google Inc.

Alex Yu Khripin

Alfred Anthony Rizzi

Stephen Berard

Eric Cary Whitman

Gina Christine Fay

Benjamin Swilling

Zachary John Jackowski

John Aaron Saunders

Michael Patrick Murphy

Steven Dickinson Potter

Christopher Everett Thorne

Matthew David Malchano

Robert Eugene Paolini

Adam Komoroski

Christopher Stathis

Navid Aghasadeghi

Max Bajracharya

Michael Scott Rose

Steve Potter

Haink Tu

A computer-implemented method includes generating a joint-torque-limit model for the articulated arm based on allowable joint torque sets corresponding to a base pose of the base. The method also include receiving a first requested joint torque set for a first arm pose of the articulated arm and determining, using the joint-torque-limit model, an optimized joint torque set corresponding to the first requested joint torque set. The method also includes receiving a second requested joint torque set for a second arm pose of the articulated arm and generating an adjusted joint torque set by adjusting the second requested joint torque set based on the optimized joint torque set. The method also includes sending the adjusted joint torque set to the articulated arm.

Limiting arm forces and torques

A method of constrained mobility mapping includes receiving from at least one sensor of a robot at least one original set of sensor data and a current set of sensor data. Here, each of the at least one original set of sensor data and the current set of sensor data corresponds to an environment about the robot. The method further includes generating a voxel map including a plurality of voxels based on the at least one original set of sensor data. The plurality of voxels includes at least one ground voxel and at least one obstacle voxel. The method also includes generating a spherical depth map based on the current set of sensor data and determining that a change has occurred to an obstacle represented by the voxel map based on a comparison between the voxel map and the spherical depth map. The method additional includes updating the voxel map to reflect the change to the obstacle.

Constrained mobility mapping

An example robot includes a hydraulic actuator cylinder controlling motion of a member of the robot. The hydraulic actuator cylinder comprises a piston, a first chamber, and a second chamber. A valve system controls hydraulic fluid flow between a hydraulic supply line of pressurized hydraulic fluid, the first and second chambers, and a return line. A controller may provide a first signal to the valve system so as to begin moving the piston based on a trajectory comprising moving in a forward direction, stopping, and moving in a reverse direction. The controller may provide a second signal to the valve system so as to cause the piston to override the trajectory as it moves in the forward direction and stop at a given position, and then provide a third signal to the valve system so as to resume moving the piston in the reverse direction based on the trajectory.

Braking and regeneration control in a legged robot

A method for calibrating a position measurement system includes receiving measurement data from the position measurement system and determining that the measurement data includes periodic distortion data. The position measurement system includes a nonius track and a master track. The method also includes modifying the measurement data by decomposing the periodic distortion data into periodic components and removing the periodic components from the measurement data.

Magnetic encoder calibration

A method of footstep contact detection includes receiving joint dynamics data for a swing phase of a swing leg of the robot, receiving odometry data indicative of a pose of the robot, determining whether an impact on the swing leg is indicative of a touchdown of the swing leg based on the joint dynamics data and an amount of completion of the swing phase, and determining when the impact on the swing leg is not indicative of the touchdown of the swing leg, a cause of the impact based on the joint dynamics data and the odometry data.

Footstep contact detection

The disclosure provides systems and methods for mitigating slip of a robot appendage. In one aspect, a method for mitigating slip of a robot appendage includes (i) receiving an input from one or more sensors, (ii) determining, based on the received input, an appendage position of the robot appendage, (iii) determining a filter position for the robot appendage, (iv) determining a distance between the appendage position and the filter position, (v) determining, based on the distance, a force to apply to the robot appendage, (vi) causing one or more actuators to apply the force to the robot appendage, (vii) determining whether the distance is greater than a threshold distance, and (viii) responsive to determining that the distance is greater than the threshold distance, the control system adjusting the filter position to a position, which is the threshold distance from the appendage position, for use in a next iteration.

Continuous slip recovery

A dynamic planning controller receives a maneuver for a robot and a current state of the robot and transforms the maneuver and the current state of the robot into a nonlinear optimization problem. The nonlinear optimization problem is configured to optimize an unknown force and an unknown position vector. At a first time instance, the controller linearizes the nonlinear optimization problem into a first linear optimization problem and determines a first solution to the first linear optimization problem using quadratic programming. At a second time instance, the controller linearizes the nonlinear optimization problem into a second linear optimization problem based on the first solution at the first time instance and determines a second solution to the second linear optimization problem based on the first solution using the quadratic programming. The controller also generates a joint command to control motion of the robot during the maneuver based on the second solution.

Dynamic planning controller

A legged robot may seek to operate according to a target gait. The legged robot may include leg members and leg joints. Possibly based on the target gait and state of the legged robot, an ordered list of gait controllers may be obtained. The gait controllers in the ordered list may define respective gaits of the legged robot, and may include respective validity checks and output parameters for the respective gaits. The ordered list may begin with a target gait controller that defines the target gait. The ordered list may be traversed in order from the target gait controller until a validity check associated with a particular gait controller passes. The legged robot may be instructed to actuate the leg members and/or leg joints according to output parameters of the particular gait controller.

Achieving a target gait behavior in a legged robot

A control system may receive a first plurality of measurements indicative of respective joint angles corresponding to a plurality of sensors connected to a robot. The robot may include a body and a plurality of jointed limbs connected to the body associated with respective properties. The control system may also receive a body orientation measurement indicative of an orientation of the body of the robot. The control system may further determine a relationship between the first plurality of measurements and the body orientation measurement based on the properties associated with the jointed limbs of the robot. Additionally, the control system may estimate an aggregate orientation of the robot based on the first plurality of measurements, the body orientation measurement, and the determined relationship. Further, the control system may provide instructions to control at least one jointed limb of the robot based on the estimated aggregate orientation of the robot.

Natural pitch and roll

An example implementation involves controlling robots with non-constant body pitch and height. The implementation involves obtaining a model of the robot that represents the robot as a first point mass rigidly coupled with a second point mass along a longitudinal axis. The implementation also involves determining a state of a first pair of legs, and determining a height of the first point mass based on the model and the state of the first pair of legs. The implementation further involves determining a first amount of vertical force for at least one leg of the first pair of legs to apply along a vertical axis against a surface while the at least one leg is in contact with the surface. Additionally, the implementation involves causing the at least one leg of the first pair of legs to begin applying the amount of vertical force against the surface.

Growing community of inventors

Mountain View, CA, United States of America

Alex Yu Khripin