The present embodiment relates to a cloud-based robot control method for controlling a plurality of robots which are positioned in a plurality of spaces divided arbitrarily, the method comprising the steps of: generating a control base model which can be applied to the plurality of robots in a cloud server; distributing the control base model to edge servers allocated to respective spaces; upgrading the control base model in accordance with the plurality of robots of a space, in the edge server; directly transmitting the upgraded control model from the edge server to another edge server; and controlling the plurality of robots by means of the upgraded control model in the edge server. Therefore, by sharing a deep-learning model among edge servers, supporting heterogeneous robots and heterogeneous services is possible. Further, a base deep-learning model from the cloud server is tuned into a customized deep-learning model to be suitable for respective robots in the edge server, and the deep-learning model is upgraded to an adaptive deep-learning model to be suitable for a service provided by respective robots, and thus an optimized service can be provided.

An assistant device according to the present disclosure includes a memory, and a hardware processor coupled to the memory. The hardware processor being configured to: add latency information indicating required latency to data that is transmitted to a server of a plurality of servers via a communication line and used for an assistance process performed by the server; transmit data for external processing to which the latency information is added, and receive assistance information including a processing result of the server based on the data for external processing; and control a vehicle based on the assistance information.

A communication control server includes circuitry to receive, from a first mobile apparatus, status information indicating a status of the first mobile apparatus currently communicating with a communication terminal as a communication destination of the communication terminal. The first mobile apparatus is movable in a real space and remotely operable by the communication terminal. The circuitry switches the communication destination of the communication terminal from the first mobile apparatus to a second mobile apparatus based on the status information. The second mobile apparatus is movable in the real space and remotely operable by the communication terminal.

A control system for controlling transport of a vehicle in any of steps from production to shipment of the vehicle includes a state identification unit that identifies a vehicle state that is a state of the vehicle, and a control notification unit that determines a control content of the vehicle by using the identified vehicle state and notifies the vehicle of the control content. The control notification unit determines to control at least either one of a transport route of the vehicle and a timing of starting the transport as the control content such that a magnitude of a difference between the vehicle state and a preset target state is suppressed.

A system for loading and transporting parcels includes: a sorter including a plurality of chutes for offloading parcels from the sorter; a plurality of totes; a plurality of self-driving vehicles (SDVs) configured to transport the plurality of totes between a loading area, an unloading area, and a queue area; and a control subsystem. The loading area includes a plurality of zones, with each zone corresponding to one or more chutes of the sorter. The control subsystem includes a controller, which is operably connected to the SDVs, and which selectively communicates instructions to dispatch SDVs to transport and replace totes in the loading area as they become filled to the predetermined capacity. A method for loading and transporting parcels in a sorting facility including a loading area, an unloading area, and a queue area is also disclosed.

A transport system includes: a plurality of transport vehicles capable of autonomously moving between a departure place and a destination; a terminal used for inputting a transport request instruction including information about a plurality of transported objects to be respectively transported by the plurality of transport vehicles; and a server capable of communicating with the plurality of transport vehicles and the terminal. The server calculates priorities of the plurality of transported objects to be transported from the departure place to the destination, based on the transport request instruction, and assigns, in accordance with the priorities, the plurality of transport vehicles to respectively transport the plurality of transported objects from transport vehicles in a state of being capable of transporting, out of the plurality of transport vehicles, and outputs, at a predetermined timing, a movement instruction to each of the assigned plurality of transport vehicles.

An operation management system including a computer configured to manage operations of registered mobilities. The computer stores road surface information in association with a position in map data based on at least a detection result of a surroundings monitoring device and positional information received from a measurement mobility. The computer stores, as a detail detection area, at least one of a poor detection area in the map data in which detection of the road surface information by the measurement mobility is poor and a predetermined area in the map data. The computer transmits, to the measurement mobility that is scheduled to travel or that is traveling the detail detection area, a speed change command, such that the measurement mobility, which travels at a normal command speed by automated driving, travels the detail detection area at a specific speed lower than the normal command speed.

A method and an apparatus for determining a position of a shelf are provided. The method may include: obtaining a number of automated guided vehicles with shelf scanning devices; determining, based on the number of the automated guided vehicles, a scanning area of a place to which each automated guided vehicle belongs; determining, based on the scanning area, a scanning route of the scanning area to which each automated guided vehicle belongs; transmitting the scanning route of the scanning area to which the automated guided vehicle belongs, to the automated guided vehicle; and determining a position of a shelf in the scanning area to which the automated guided vehicle belongs based on scanning information of a shelf scanning device on the automated guided vehicle and position information of the automated guided vehicle.

The present application provides an unmanned aerial vehicle (UAV)-assisted federated learning resource allocation method for an UAV-assisted federated learning wireless network scenario, which takes into account the effect of altitude of the UAV on the coverage range in order to achieve an equilibrium between the total energy consumption of the user and federated learning performance. The method simultaneously considers the total energy consumption of the user and the federated learning performance, defines the total cost function of the system. The total cost function consists of weighting of the total energy consumption of the user and the inverse of the number of users participating in federated learning, and forms the optimization problem with a minimization of the total cost function.

This disclosure relates generally to method and system for multi-object tracking and navigation without pre-sequencing. Multi-object navigation is an embodied Al task where object navigation only searches for an instance of at least one target object where a robot localizes an instance to locate target objects associated with an environment. The method of the present disclosure employs a deep reinforcement learning (DRL) based framework for sequence agnostic multi-object navigation. The robot receives from an actor critic network a deterministic local policy to compute a low-level navigational action to navigate along a shortest path calculated from a current location of the robot to the long-term goal to reach the target object. Here, a deep reinforcement learning network is trained to assign the robot with a computed reward function when the navigational action is performed by the robot to reach an instance of the plurality of target objects.