An autonomous sweeper is provided, including a sweeper module, and a robot chassis having a length along a pair of sides, a front side, a back side and a top side that define an interior space. The sweeper module is configured to fit within the interior space when the robot chassis moves over the sweeper module. A pair of wheels is disposed proximate to the back side of the robot chassis and a single wheel is disposed proximate to the front side. A pair of scissor lifts is disposed along said pair of sides. A lift frame including alignment pegs that fit into corresponding alignment holes is disposed on the top side of the sweeper module. The lift frame is raised and lowered by said pair of scissor lifts, and said scissor lifts assist in lifting the sweeper module while aligning said sweeper module to the robot chassis using said alignment pegs and alignment holes.

In various examples, a machine may generate, using sensor data capturing one or more views of an environment, a virtual environment including a 3D representation of the environment. The machine may render, using one or more virtual sensors in the virtual environment, one or more images of the 3D representation of the environment. The machine may apply the one or more images to one or more machine learning models (MLMs) trained to generate one or more predictions corresponding to the environment. The machine may perform one or more control operations based at least on the one or more predictions generated using the one or more MLMs.

A method comprising applying an object recognition pipeline to frames of video data. The object recognition pipeline provides a mask output of objects detected in the frames. The method includes fusing the mask output of the object recognition pipeline with depth data associated with the frames of video data to generate a map of object instances, including projecting the mask output to a model space for the map of object instances using a camera pose estimate and the depth data. An object instance in the map of object instances is defined using surface-distance metric values within a three-dimensional object volume, and has an object pose estimate indicating a transformation of the object instance to the model space. The object pose estimate and the camera pose estimate form nodes of a pose graph for the map of model instances.

A mobile robot is configured to navigate on a sidewalk and deliver a delivery to a predetermined location. The robot has a body and an enclosed space within the body for storing the delivery during transit. At least two cameras are mounted on the robot body and are adapted to take visual images of an operating area. A processing component is adapted to extract straight lines from the visual images taken by the cameras and generate map data based at least partially on the images. A communication component is adapted to send and receive image and/or map data. A mapping system includes at least two such mobile robots, with the communication component of each robot adapted to send and receive image data and/or map data to the other robot. A method involves operating such a mobile robot in an area of interest in which deliveries are to be made.

A vehicular control system includes a forward-viewing camera, a forward-sensing sensor and a cabin-sensing imaging sensor. With the system controlling driving of the vehicle, the system determines a triggering event that triggers handing over driving of the vehicle to a driver of the vehicle before the vehicle encounters an event point associated with the triggering event. The vehicular control system (i) determines a total action time available before the vehicle encounters the event point, (ii) estimates a driver takeover time for the driver to take over control of the vehicle and (iii) estimates a handling time for the driver to control the vehicle to avoid encountering the event point. The vehicular control system estimates the driver takeover time based at least in part on processing by the vehicular control system of the in-cabin image data captured by the cabin-sensing imaging sensor.

The assisted navigation system is intended to enable an assisted operation mode in ground mobile robots. The system is designed to achieve an autonomous relocation of a robot from one location to another location within a sidewalk, minimizing the need for constant human intervention. The system includes a camera for collecting the visual data needed to assess the terrain and potential obstacles, a collection of sensors to detect potential obstacles during assisted operations, a communication module to receive inputs from a remote operator to enable the activation of this system, a localization module for teleoperations, a local server module to store the information gathered, a processor configured to operate a robot in an assisted mode of operation based on input from the communication interface in which the robot performs a task without human intervention, and a communication interface coupled to the processor and configured to communicate control values to the systems of the mobile robot.

A system and method involves detecting a trigger event during operation of an autonomous ground vehicle traveling between two physical locations; generating event sequence data from primary sensor data, secondary sensor data, spatiotemporal data, and telemetry data through operation of a reporter; communicating the event sequence data to cloud storage and raw data to a streaming database; transforming the raw data into normalized data stored in a relational database through operation of a normalizer; operating a curation system to identify true trigger events from the normalized data and extract training data by way of a discriminator; operating a machine learning model within an active learning pipeline to generate a model update from aggregate training data generated from the training data by an aggregator; and reconfiguring the navigational control system with the model update communicated from the active learning pipeline to the autonomous ground vehicle.

A three-layer intelligence system architecture and an exploration robot are provided. The three-layer intelligence system architecture includes: a digital twin module for creating a virtual exploration environment and a virtual robot according to explored environment data acquired in real time by the exploration robot and robot data of the exploration robot; a virtual reality module for generating a process and a result of the virtual robot executing the control commands in the virtual exploration environment according to the virtual exploration environment, the virtual robot, and control commands of a control personnel for the exploration robot; and a man-machine fusion module for transmitting the control commands and showing the control personnel the process and the result of the virtual robot executing the control commands in the virtual exploration environment, and causing the exploration robot to execute the control commands after acquiring a feedback indicating that the control personnel confirms the control commands.

A method for a stair tracking for modeled and perceived terrain includes receiving, at data processing hardware, sensor data about an environment of a robot. The method also includes generating, by the data processing hardware, a set of maps based on voxels corresponding to the received sensor data. The set of maps includes a ground height map and a map of movement limitations for the robot. The map of movement limitations identifies illegal regions within the environment that the robot should avoid entering. The method further includes generating a stair model for a set of stairs within the environment based on the sensor data, merging the stair model and the map of movement limitations to generate an enhanced stair map, and controlling the robot based on the enhanced stair map or the ground height map to traverse the environment.

A conveyor station, robot module, sweeper module, and methods for autonomously emptying debris using the conveyor station are described. In one example, a conveyor station includes a housing having an input end and an output end. The conveyor station includes a conveyor belt having a receiving region proximate to the input end and an angled transport region leading toward a dispense region. The conveyor belt has a plurality of fins that extend out from a surface of the conveyor belt. The plurality of fins enable movement of debris collected at the receiving region toward the dispense region. The dispense region is configured to push debris into a drop funnel of the housing, and the drop funnel directs debris into a receptacle. The conveyor station includes a conveyor controller of the conveyor station is configured with a sensor for detecting presence of a sweeper module. The sweeper module includes a container that holds debris collected when the sweeper module is connected to a robot module. The debris is configured to be emptied from said sweeper module directly onto said receiving region of the conveyor belt.