Abstract: A method for controlling a robotic item handler is described. The method includes constructing a machine learning model based on a combined point cloud data as an input to a convolution neural network, outputting a decision classification indicative of a first probability associated with a first operating mode and a second probability associated with a second operating mode, selecting one of the first operating mode and the second operating mode of the robotic item handler based on the decision classification outputted by the machine learning model, and operating the robotic item handler according to the selection of the one of the first operating mode and the second operating mode.

Abstract: A mobility platform is configured to execute one or more tasks in a worksite including a passive landmark. A mobility platform may include a first laser rangefinder and at least one processor configured to sweep the first passive landmark with the first laser rangefinder to collect a first plurality of distance measurements for a first plurality of yaw angles, fit a first shape to the first plurality of distance measurements based on a predetermined shape of the first passive landmark, and determine a position of a geometric center of the first passive landmark relative to the first location of the first laser rangefinder based on the fit first shape.

Abstract: Various embodiments are directed to unloading system and methods of using the same. In various embodiments, an unloading system comprises a trailer bed configured to receive a container thereon, wherein the trailer bed comprises one or more joints configurable between a plurality of joint positions, wherein the trailer bed is configured to lift, tilt, and oscillate at least a portion of the container based on the joints; and an extendable conveyor configured to receive a box that has been released from the container and convey the box away from the trailer bed. Various embodiments are directed to a method comprising tilting the container such that an open end of the container is lower than a closed end of the container, wherein the open end of the container is proximate to an extendable conveyor; and oscillating the container to facilitate release of a box from the container onto the extendable conveyor.

Abstract: A method for controlling a robotic item handler is described. The method includes obtaining first point cloud data related to a first three-dimensional (3D) image and second point cloud data related to a second 3D image, captured by a first sensor device and a second sensor device respectively. Further, the method can include transforming the first point cloud data and the second point cloud data to combined point cloud data that is used as an input to a convolutional neural network, to construct a machine learning model. The machine learning model can output a decision classification indicative of a first probability associated with a first operating mode and a second probability associated with a second operating mode. Furthermore, the method can include operating the robotic item handler according to the first operating mode or the second operating mode based on a comparison of the first probability and the second probability.

Abstract: A mobility platform is configured to execute one or more tasks in a worksite including a first passive landmark and a second passive landmark. The mobility platform may include a chassis, a drive system supporting the chassis, a first laser rangefinder disposed on the chassis at a first location, a second laser rangefinder disposed on the chassis at a second location, and at least one processor. The at least one processor may be configured to determine a position and orientation of the chassis based on a first distance measured by the first laser rangefinder between the first location and a first known landmark position, a second distance measured by the second laser rangefinder between the second location and a second known landmark position, and yaw angle information from at least one of the first and second laser rangefinders.

Abstract: Various embodiments described herein relate to techniques for reinforcement learning based conveyoring control. In this regard, a conveyor system is configured to transport one or more objects via a conveyor belt. Furthermore, a vision system comprises one or more sensors configured to scan the one or more objects associated with the conveyor system. A processing device is configured to employ a machine learning model to determine object pose data associated with the one or more objects. The processing device is further configured to generate speed control data for the conveyor belt of the conveyor system based on a set of control policies associated with the object pose data.

Abstract: A truck unloader for unloading cartons from a carton pile includes a controller with a detection system configured to periodically receive one or more images from a vision system and a processor subsystem in connection with the detection system. The processor subsystem configured to calculate a distance from at least a portion of the truck unloader to the carton pile based on the one or more images and advance the truck unloader to perform a sweeping operation based on the distance in an instance in which the distance satisfies a predefined distance threshold. The processor subsystem further configured to monitor a torque input received as a feedback from the one or more sensors while performing the sweeping operation, wherein the sweeping operation is configured to continue until the torque input exceeds a predefined torque threshold.

Abstract: A method for controlling a robotic item handler is described. The method includes obtaining first point cloud data related to a first three-dimensional (3D) image and second point cloud data related to a second 3D image, captured by a first sensor device and a second sensor device respectively. Further, the method can include transforming the first point cloud data and the second point cloud data to combined point cloud data that is used as an input to a convolutional neural network, to construct a machine learning model. The machine learning model can output a decision classification indicative of a first probability associated with a first operating mode and a second probability associated with a second operating mode. Furthermore, the method can include operating the robotic item handler according to the first operating mode or the second operating mode based on a comparison of the first probability and the second probability.

Abstract: Various embodiments are directed to unloading system and methods of using the same. In various embodiments, an unloading system comprises a trailer bed configured to receive a container thereon, wherein the trailer bed comprises one or more joints configurable between a plurality of joint positions, wherein the trailer bed is configured to lift, tilt, and oscillate at least a portion of the container based on the joints; and an extendable conveyor configured to receive a box that has been released from the container and convey the box away from the trailer bed. Various embodiments are directed to a method comprising tilting the container such that an open end of the container is lower than a closed end of the container, wherein the open end of the container is proximate to an extendable conveyor; and oscillating the container to facilitate release of a box from the container onto the extendable conveyor.

Abstract: Various embodiments described herein relate to techniques for reinforcement learning based conveyoring control. In this regard, a conveyor system is configured to transport one or more objects via a conveyor belt. Furthermore, a vision system comprises one or more sensors configured to scan the one or more objects associated with the conveyor system. A processing device is configured to employ a machine learning model to determine object pose data associated with the one or more objects. The processing device is further configured to generate speed control data for the conveyor belt of the conveyor system based on a set of control policies associated with the object pose data.