A method of generating driving routes of a plurality of mobile robots including generating a plurality of virtual driving lines in a driving space, generating a pattern of driving behaviors of the mobile robots based on the plurality of generated virtual driving lines, inputting an initial position and a final position of each of the plurality of mobile robots, and generating respective driving routes from the initial position to the final position of each of the plurality of mobile robots on the plurality of generated virtual driving lines based on the generated pattern of the driving behaviors.

A method of loading and/or unloading a loading space is provided having a loading vehicle that drives into the loading space at least once to place down and/or to collect at least one load object, wherein an access zone of the loading space is safeguarded by at least one first sensor and the loading vehicle is safeguarded by at least one second sensor, In this respect, on driving into the loading space, the loading vehicle first drives to a first position that is so close to the safeguarded access zone that no person fits between the loading vehicle and the safeguarded access zone and the safeguarding of the access zone is then adapted by the first sensor such that a drive-through corridor for the loading vehicle is created.

An information processing apparatus acquires a route plan of a movable apparatus; acquires information regarding an obstacle factor hindering a way of the route plan of the movable apparatus; calculates the degree of obstacle for the route plan of the movable apparatus based on the information regarding the obstacle factor; determines an obstacle countermeasure for reducing the degree of obstacle based on the information regarding the obstacle factor and the route plan; and performs at least one of control for causing the obstacle factor evacuate and control for the movable apparatus to avoid the obstacle factor based on the obstacle countermeasure.

An information processing apparatus acquires a route plan of a movable apparatus; acquires information regarding an obstacle factor hindering a way of the route plan of the movable apparatus; calculates the degree of obstacle for the route plan of the movable apparatus based on the information regarding the obstacle factor; determines an obstacle countermeasure for reducing the degree of obstacle based on the information regarding the obstacle factor and the route plan; and performs at least one of control for causing the obstacle factor evacuate and control for the movable apparatus to avoid the obstacle factor based on the obstacle countermeasure.

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.

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.

The invention relates to method (100) for detecting at least one obstacle in an automated and/or at least semi-autonomous driving system (60), said method comprising the following steps: providing (101) image data, wherein the image data are specific to a recording of an environment of the driving system (60), performing (102) an evaluation of the image data provided, wherein the evaluation takes place based on an application of a machine learning model (50), by means of which an occlusion label is determined for at least one occlusion of the environment, performing (103) the detection of the at least one obstacle on the basis of the occlusion label determined.

The invention relates to method (100) for detecting at least one obstacle in an automated and/or at least semi-autonomous driving system (60), said method comprising the following steps: providing (101) image data, wherein the image data are specific to a recording of an environment of the driving system (60), performing (102) an evaluation of the image data provided, wherein the evaluation takes place based on an application of a machine learning model (50), by means of which an occlusion label is determined for at least one occlusion of the environment, performing (103) the detection of the at least one obstacle on the basis of the occlusion label determined.

The present invention relates to a road crossing method for a mobile robot. The road crossing method comprises the mobile robot approaching a road crossing. Further, the road crossing method comprises estimating, with a data processing unit, a location and time of collision with at least one dynamic object on the road crossing. Further still, the road crossing method comprises generating, with the data processing unit, control commands for the mobile robot to avoid collision with the at least one dynamic object based on the estimated location and time of collision with the at least one dynamic object. In addition, the present invention relates to a mobile robot comprising the data processing unit and configured to carry out the road crossing method. In a further aspect, the present invention relates to a positioning method for a wheeled mobile robot positioned on a sloped terrain, comprising the mobile robot performing at least one maneuver for minimizing a magnitude of an acceleration vector of the mobile robot due to the gravity force acting on the mobile robot. In addition, the present invention relates to a mobile robot configured to carry out the positioning method.

An unmanned vehicle includes: a travel device; an obstacle sensor; a host path storage unit that stores a host path; a travel control unit that controls the travel device based on the host path; an oncoming path storage unit that stores an oncoming path to be given to an oncoming vehicle; and an obstacle presence/absence determination unit that determines whether or not an obstacle is located on the oncoming path based on detection data from the obstacle sensor.