Systems and methods for generating models for automated handling of vehicles
AssigneeAmazon Technologies, Inc.
InventorsDavid L. Alspaugh, Dipan M. Shah, Pragyana K. Mishra
Life of this patent
Estimating the term…
Abstract
Systems and methods relating to generations of models to facilitate safe, automated handling and maneuvering of vehicles, such as unmanned aerial vehicles (UAV), by robotic systems, such as a robotic arm. The described systems and methods can include a robotic system, such as a robotic arm, having a load cell to measure certain forces and torques to generate models representing the behavior of vehicles and surfaces on which the vehicles may be placed and/or from which the vehicles may be removed.
Claims20
A computer-implemented method, comprising:
- causing a first robotic arm to initiate placement of a first aerial vehicle on a first surface;
- as the first aerial vehicle is being placed on the first surface, measuring, using a load cell, at least one of a force or a torque exerted on the first aerial vehicle;
- generating a model representing an interaction between the first aerial vehicle and the first surface based at least in part on at least one of the force or the torque, the model corresponding to a first aerial vehicle type associated with the first aerial vehicle and a first surface type associated with the first surface;
- storing the model in a datastore;
- determining, for a second aerial vehicle that is to be placed on a second surface, that a second aerial vehicle type associated with the second aerial vehicle is the same as the first aerial vehicle type;
- determining that a second surface type associated with the second surface is the same as the first surface type;
- obtaining the model from the datastore;
- and
- applying the model in causing at least one of the first robotic arm or a second robotic arm to place the second aerial vehicle on the second surface.
A system, comprising:
- a robotic arm;
- a datastore;
- and
- a control system including one or more processors and program instructions that, when executed by the one or more processors, cause the one or more processors to at least:
- cause the robotic arm to initiate placement of a vehicle engaged with the robotic arm on a surface;
- measure at least one of a force or a torque exerted on the vehicle as the vehicle is placed on the surface;
- generate a model, based at least in part on at least one of the force or the torque, representing an interaction of the vehicle with the surface;
- and
- store the model in the datastore.
A robotic arm, comprising:
- an end effector;
- and
- a control system including one or more processors and program instructions that, when executed by the one or more processors, cause the one or more processors to at least:
- determine a vehicle type associated with a vehicle engaged by the end effector;
- determine a surface type associated with a surface;
- obtain a model corresponding to the vehicle type and the surface type;
- and
- apply the model in causing the robotic arm to at least one of:
- place the vehicle on the surface, or
- raise the vehicle from the surface.
Dependencies are inferred from the claim text.