Method and system for navigating mobile logistics robots

EP4457572A1Pending Publication Date: 2024-11-06STILL GMBH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
EP2022823378
Authority / Receiving Office
EP · EP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-12-29
Filing Date
2022-11-29
Publication Date
2024-11-06

AI Technical Summary

Technical Problem

The challenge in navigating mobile logistics robots indoors lies in the inaccuracies and inflexibility of metric maps, which are either self-created with high technological risks or third-party mapped, and the unsuitability of topological maps for navigation without changes, especially due to odometry errors and hardware deviations.

Method used

A smart topological map is created without global metric map information, using uniquely designed landmarks detected by sensors, which are recognized and stored as nodes, allowing for reliable navigation by decoupling global navigation from the metric environment and enabling the sharing of maps between robots.

Benefits of technology

This approach enables reliable navigation of mobile logistics robots indoors without relying on global metric maps, allowing for agile adaptation to changes and efficient path planning, reducing computational load and enabling localization without GPS or similar systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 1.1
    Figure 1.1
Patent Text Reader

Abstract

The invention relates to a method for navigating mobile logistics robots (1) by means of a map which is created by detecting a working environment (8) of the mobile logistics robot (1) by means of a sensor system. According to the invention, the map is created as an intelligent topological map without using global metric map information, wherein landmarks (2), which are uniquely designed by an identification system (7), in the working environment (8) of the mobile logistics robot (1) are detected by means of the sensor system, identified as such by analysis of the sensor data in a data processing unit, and stored as nodes in the topological map. The invention also relates to a system for carrying out the method.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Description

[0002] Method and system for navigation of mobile logistics robots

[0003] The invention relates to a method for navigating mobile logistics robots using a map that is created by detecting a working environment of the mobile logistics robot using sensors.

[0004] The invention also relates to a system for carrying out the method.

[0005] Mobile robots are increasingly being used in industry and logistics companies to automate processes in industrial manufacturing and logistics tasks, such as order picking. Mobile robots with arm manipulators, especially robot arms, are typically used. One example of this is so-called articulated-arm robots.

[0006] The implementation of mobile logistics robots, especially autonomous industrial trucks with robotic arms for load handling, such as mobile order-picking robots, is particularly challenging because they are expected to move freely within a logistics area, such as a warehouse. In doing so, mobile logistics robots constantly encounter completely new working environments.

[0007] Localization is a fundamental requirement for the navigation of automated vehicles in intralogistics environments. Localization is based on matching an expected value with the sensory perception of the vehicle. If the sensory perception successfully matches the expected value, the vehicle's position can be calculated. A distinction is made between relative localization and global localization.

[0008] To implement self-driving vehicles, both localization and vehicle guidance are necessary. A closed or open control loop calculates suitable target values ​​for vehicle propulsion based on a target description and the localization, and adjusts them if necessary. For global navigation—that is, a combination of localization and vehicle guidance—a map is often used in addition to localization. This map contains connection information in the form of obstacles, graphs, or routes. Maps are therefore used to calculate a suitable connection based on connection information to connect the vehicle's current position with the target position.

[0009] Numerous types of maps exist, including metric, semantic, or topological maps, for depicting connectivity information. The effort required to create maps varies greatly and requires different levels of complexity.

[0010] The creation of metric maps for navigation today represents the standard for the navigation of automated vehicles in the intralogistics environment. The challenge is that the creation of metric map material is associated either with high technological risks, as is particularly the case with self-mapping, or with complex surveying of the movement area, as is particularly the case with external mapping.

[0011] With self-mapping, the vehicle creates its own map of its surroundings through exploratory drives. Localizing automated vehicles indoors is particularly challenging, as satellite positioning is not possible for physical reasons. For the vehicle to create a metric map of its surroundings, it must specifically drive through the unknown areas, such as a warehouse, and map them in the process. Sensor inaccuracies (random errors) and calibration differences (systematic errors) result in different metric maps for each vehicle, each with integral errors. Vehicles that want to use these maps must develop technological solutions to address these inaccuracies.

[0012] Maps created by external mapping have little or no random and systematic errors. However, these errors are difficult to adapt to environmental changes, are inflexible, and generally expensive. Topological maps are a type of map designed to store connection information of the movement space. In contrast to metric maps, topological maps are a very simple type of map in which the movement space is represented by edges and nodes. The disadvantage of these maps is that they are not suitable for navigation without modifications.

[0013] The use of metric maps is robot-specific due to odometry errors and variations in hardware components, such as laser scanners. Providing the robot's own metric map to other mobile robots is difficult due to the resulting differences between the detected environment and the stored metric map.

[0014] The present invention is based on the object of designing a method of the type mentioned and a system for carrying out the method in such a way that reliable navigation of the mobile logistics robots is also possible indoors.

[0015] This object is achieved according to the invention in that the map is created as an intelligent topological map without using global metric map information, whereby landmarks in the working environment of the mobile logistics robot, which are uniquely designed via an identification system, are detected by means of the sensors, recognized as such by evaluating the sensor data in a data processing unit and stored as nodes in the topological map.

[0016] To decouple the global navigation of mobile logistics robots from the metric environment, the invention introduces a smart topological map, i.e., an intelligent topological map. This smart topological map consists of nodes that describe unique landmarks.

[0017] Preferably, real markers are used as landmarks, which are placed at prominent positions in the working environment of the mobile logistics robot, in particular at intersections and / or curves and / or stations and / or elevators and / or gates. The identification system preferably comprises number and / or letter codes and / or pictograms, which can be perceived by the sensors, expediently attached to the mobile logistics robot, and recognized by evaluating the sensor data in the data processing unit, expediently also housed on the mobile logistics robot, and used for automatic navigation of the mobile logistics robot in a vehicle control unit.

[0018] The mobile logistics robot can therefore detect these unique landmarks using sensors, recognize them as such, and store them as nodes in the topological map.

[0019] The nodes of the topological map are connected by edges. These edges can indicate the direction of movement of the connection between the individual nodes, particularly through a predecessor / successor principle.

[0020] According to a particularly preferred embodiment of the invention, these edges connecting the nodes are stored in the topological map with additional annotations of the edges.

[0021] The annotations of the edges advantageously contain information about a preferred path between two nodes and / or about a distance between two nodes and / or about a road width and / or about road obstructions, for example, the number of people, oncoming traffic, blockages, and / or about speed restrictions and / or about regions with interaction options.

[0022] The mobile logistics robot navigates along the nodes or the acquired or loaded node network. A possible navigation path for the mobile logistics robot is then conveniently determined by determining a valid node combination between a starting point and a destination point on the intelligent topological map using the data processing unit. The starting point can be found using vehicle behavior heuristics. For example, the mobile logistics robot could be instructed to follow an obstacle-oriented path to the next node. The task of the mobile logistics robot for executing the movement is therefore no longer to navigate along a global route, but only to the next node. All the information required for this is contained in the annotated edge information of the connecting element, i.e. the edge to specific nodes.This information is also subject to systematic random errors, but it is much more reliable because the movement space to the next node is much smaller, making integral errors less significant. Because the topological map according to the invention is completely decoupled from the metric environment after recording, but global references are created at important forks in the path and recognized by the mobile robot, navigation can occur without a global metric map, which is considered a significant innovation.

[0023] The mobile logistics robot's navigation between the nodes is advantageously carried out reactively, with the mobile logistics robot orienting itself to the existing environment through environmental perception using its sensors. The mobile logistics robot can derive its driving behavior from the existing environment, be it obstacles, lane markings, or other orientation features.

[0024] According to a preferred development of the inventive concept, the mobile logistics robot shares the created intelligent topological map with at least one other mobile logistics robot. The other mobile logistics robot can then compare and supplement the received topological map with its own map. This creates an overview of the environment in which the connections between the individual landmarks are known. Using the smart topological map, paths and tasks can be planned along these connections between the nodes. The planned paths are then processed along the known landmarks and confirmed by recognizing the landmarks along the route.

[0025] Local navigation, obstacle detection and avoidance, interaction with objects and stations can be handled by the on-board sensors and systems and only take place locally.

[0026] If paths change, for example, due to a permanent blockage of a connection between two nodes, the removal or addition of nodes, etc., the topological map can be agilely updated. By sharing the smart topological map, other mobile robots are directly informed of the changes and can thus adapt to new situations.

[0027] In a further advantageous embodiment of the invention, at least two mobile logistics robots share information about their planned paths in the intelligent topological map, with this information being used in at least one traffic management system to control traffic situations. For example, future nodes can only be released if they are not occupied by vehicles ahead.

[0028] Another development of the invention provides that the mobile logistics robot shares the created intelligent topological map with at least one manually or semi-automatically controlled vehicle to enable localization of the manually or semi-automatically controlled vehicle. In this way, the smart topological maps can also be used for manually or semi-automated vehicles and their localization.

[0029] In a further embodiment, it is possible to generate a task planning based on the smart topological map, so that the vehicles create processing hierarchies that can be used for individual transports.

[0030] The invention further relates to a system for carrying out the method with at least one mobile logistics robot with a sensor system for detecting a working environment of the mobile logistics robot and a data processing unit for evaluating the sensor data.

[0031] The system solves the task by providing landmarks in the working environment of the mobile logistics robot that are uniquely designed using an identification system and can be detected by the sensors. The data processing unit is configured to recognize the landmarks as such by evaluating the sensor data and to store them as nodes in the topological map. 1

[0032] The identification system expediently comprises optical markers, while the sensor system comprises at least one optical sensor, in particular a camera, which is preferably attached to the mobile logistics robot and serves for the necessary sensory perception.

[0033] Radio-based node detection is also conceivable. In this case, the identification system comprises radio-based transmitters at the landmarks, while the sensor system includes at least one radio-based receiver.

[0034] The invention offers a whole range of advantages:

[0035] The invention enables navigation of automatic vehicles without global metric maps.

[0036] A map of a mobile robot's workspace can be shared with other mobile robots. If one mobile robot detects changes in the area, other mobile robots can learn from it.

[0037] Multi-story workspaces can be easily managed without any additional software thanks to the smart topological map. It enables automated vehicles to navigate and locate themselves using elevators in multi-story warehouses, for example.

[0038] The localization of mobile robots is no longer dependent on a global or local positioning system, such as GPS, Galileo or WLAN.

[0039] Heuristic algorithms can use the smart topological map to plan paths and tasks. The smart topological map can provide heuristic weighting parameters, such as distance, lane width, curve frequency, pedestrian traffic, and speed limits. Because the intelligent topological map compactly bundles nodes and edges, global, metric occupancy grid maps do not need to be processed. This saves computing time. Through an expanded exchange of their planned paths, mobile robots can detect whether other participants also have a planned path on the edges used, thus recognizing emerging traffic situations and responding to them with traffic management.

[0040] Further advantages and details of the invention are explained in more detail with reference to the exemplary embodiments shown in the schematic figures.

[0041] Figure 1 shows a landmark in a warehouse from the perspective of a mobile logistics robot and

[0042] Figure 2 shows the situation from Figure 1 with the mobile logistics robot from a perspective view from above.

[0043] Figure 1 shows an example of a landmark 2 in a warehouse 5. Figure 1 shows only one landmark 2, but typically there are several landmarks 2 in the warehouse 5. The landmarks 2 can vary by any desired unique identifier.

[0044] The landmarks 2 are physically placed markers that are uniquely designed using an identification system 7, for example, numbers, letter codes, or pictograms, and can be perceived sensorially by the mobile logistics robot 1, which can be seen in the illustration in Figure 2. In the case shown, the landmark 2 is a sign 6 with unique ArUco markers 3 as machine-readable characters. The landmark 2 also contains a representation 4 for human interaction as a translation of the machine-readable characters.

[0045] The landmarks 2 are placed at prominent positions, such as intersections, curves, stations, elevators, and gates, in the accessible robot area, i.e., the working environment 8 of the mobile logistics robot 1. The mobile logistics robot can detect these unique landmarks 2 using its on-board sensors, recognize them as such, and store them as nodes in a topological map. To decouple the global navigation of the mobile logistics robot 1 from the metric environment, a smart topological map, i.e., an intelligent topological map, is used. This smart topological map consists of nodes formed by the unique landmarks 2.

[0046] Figure 2 shows the situation from Figure 1 with the mobile logistics robot from a perspective view from above. This illustration also includes the mobile logistics robot 1. Furthermore, several unique landmarks 2 in the form of signs 6 can be seen, which are attached at prominent positions in the warehouse 5. The mobile logistics robot 1 can detect these unique landmarks 2 using sensors, recognize them as such, and store them as nodes in the topological map. The navigation of the mobile logistics robot 1 occurs along the nodes that describe the landmarks 2 in the topological map, or rather, along the recorded or loaded node network.

Claims

Patent claims Method for navigating mobile logistics robots (1) using a map created by capturing a working environment (8) of the mobile logistics robot (1) using sensors, characterized in that the map is created as an intelligent topological map without using global metric map information, wherein landmarks (2) in the working environment (8) of the mobile logistics robot (1), which are uniquely designed via an identification system (7), are captured by means of sensors, recognized as such by evaluating the sensor data in a data processing unit and stored as nodes in the topological map.A method according to claim 1, characterized in that the landmarks (2) are real, placed markers located at prominent positions in the working environment (8) of the mobile logistics robot (1), in particular at intersections and / or curves and / or stations and / or elevators and / or gates. A method according to claim 1 or 2, characterized in that the identification system (7) comprises numerical and / or letter codes and / or pictograms. A method according to any one of claims 1 to 3, characterized in that edges connecting the nodes are stored in the topological map with additional edge annotations.The method according to claim 4, characterized in that the edge annotations contain information about a preferred path between two nodes and / or about a distance between two nodes and / or about a lane width and / or about lane obstructions and / or about speed limits and / or about regions with interaction possibilities. A method according to any one of claims 1 to 5, characterized in that the navigation of the mobile logistics robot (1) is carried out by determining a valid node combination between a starting point and a destination point on the intelligent topological map using the data processing unit. A method according to any one of claims 1 to 6, characterized in that the navigation of the mobile logistics robot (1) between the nodes is carried out reactively, wherein the mobile logistics robot (1) orients itself by perceiving its environment using the sensors. A method according to any one of claims 1 to 7, characterized in that the mobile logistics robot (1) shares the created intelligent topological map with at least one other mobile logistics robot (1).A method according to any one of claims 1 to 8, characterized in that at least two mobile logistics robots (1) share information about their planned paths in the intelligent topological map, wherein this information is used in at least one traffic management system for controlling traffic situations. A method according to any one of claims 1 to 9, characterized in that the mobile logistics robot (1) shares the created intelligent topological map with at least one manually or semi-automatically controlled vehicle to enable localization of the manually or semi-automatically controlled vehicle.System for carrying out the method according to one of claims 1 to 10 with at least one mobile logistics robot (1 ) with sensors for recording a working environment of the mobile logistics robot (1 ) and a data processing unit for evaluating the sensor data, characterized in that landmarks (2) are provided in the working environment (8) of the mobile logistics robot (1 ) which are uniquely designed via an identification system (7) and can be recorded by means of the sensors, and the data processing unit is configured to evaluate the. Sensor data to recognize the landmarks (2) as such and to store them as nodes in the topological map. System according to claim 11, characterized in that the identification system (7) comprises optical markers and the sensor system comprises at least one optical sensor, in particular a camera. System according to claim 11 or 12, characterized in that the identification system (7) comprises radio-based transmitters and the sensor system comprises at least one radio-based receiver.