Control network for mobile robots

Abstract

A method (300) for supporting one or more mobile robots operable in a facility equipped with two or more access networks, where the method is implemented in a queue management system, FMS, and the method comprises: assigning (310) at least one connection priority to a particular robot or robot task; performing a plan for the one or more mobile robots, the plan including a path plan (314); generating a predictive network resource request (322) based on the planned output, and sharing it with a network management system (NMS); causing the NMS to allocate (324) network resources in the access network to the one or more mobile robots (324) according to the connection priority; and executing (328) an output of the plan, the plan including the path plan.

Description

Technical Field

[0001] This disclosure relates to the field of mobile robot control. In particular, it discloses a control network for supporting one or more mobile robots operating in a facility, wherein the control network includes a fleet management system and a network management system. Background Technology

[0002] Edge computing nodes are increasingly supporting motion planning for mobile robots and mobile robot manipulators, and the corresponding robot control algorithms are typically executed within these edge computing nodes. For example, the applicant's previous publication, WO2022214193, relates to a control network for supporting one or more mobile robots operable in facilities such as factories, warehouses, ports, or container terminals. In this control network, a queue management system (FMS) is authorized to perform path planning and path execution for the mobile robots, while a network management system (NMS) is authorized to configure and perform resource allocation within a single access network. The respective authority of the FMS and NMS is mutually exclusive. The access network providing wireless connectivity for the mobile robots in the facility can be a general-purpose network, such as 5G (3GPP NR) or Wi-Fi 6 (IEEE 802.11ax), or it can use dedicated industrial wireless technologies such as WIA-FA (Wireless Network for Industrial Automation - Factory Automation, as specified in standards such as IEC PAS 62948).

[0003] To give a few more examples, in the research paper by Ortiz et al., "Fleet management system for mobile robots in healthcare environments" ( Journal of Industrial Engineering and Management Volume 14 (2021), pp. 55-71 [DOI:10.3926 / jiem.3284] discusses how to design and implement FMS to support mobile robot platoons for indoor logistics applications in medical and commercial settings.

[0004] In the paper "Constructing areliable and fast recoverable network for drones" (IEEE International Conference on Communications (ICC), 2016, pp. 1-6 [DOI:10.1109 / ICC.2016.7511317]) by Lee et al., a centralized routing protocol for multiple drones forming a mobile ad hoc network is proposed. Drones communicate with each other in a multi-hop manner to convey their control commands, so failure of any network link can have serious consequences. The routing utilizes relevant information about the drones, such as their geographical location, flight schedule, and status. A queue management module plans and creates flight control commands for each drone while monitoring their location. Information about the required routes between drones is then configured in the drones, which monitor their communication links with neighbors and report to the centralized routing when a link fails due to, for example, excessive distance.

[0005] Furthermore, Lyczkowski et al., in their paper "SDN Controlled Visible Light Communication Clusters for AGVs" presented at the 2021 European Conference on Networks and Communications and the 6G Summit (EuCNC / 6GSummit), pp. 154-159 [DOI:10.1109 / EuCNC / 6GSummit51104.2021.9482417], describe a method to improve the communication latency and reliability of AGV collaborative tasks. This method envisions using infrastructure-based links (i.e., Wi-Fi) and self-organizing networks (based on the principles of visible light communication (VLC)) to exchange control information among AGVs sharing a task. AGV clustering is performed using information about factory production and network communication status to minimize the number of handovers between Wi-Fi and VLC communication.

[0006] Ideally, the access network in a mobile robot system should be able to handle all anticipated communication scenarios with sufficient performance. However, the radio resources of each access network are inherently limited. Communication performance is particularly challenging when a swarm of mobile robots shares a wireless network with factory machinery and / or other wireless devices. Then, the performance of the mobile robots may degrade as other machines, wearable devices used by human workers, and / or handheld devices compete for the same network resources. If a large number of wireless clients continuously connect to a shared network, not only recently added clients but also existing clients may experience negative performance impacts. These negative performance impacts can include longer communication latency, data loss, or even complete refusal to connect to the network.

[0007] While the number of wireless clients allowed to connect to the network can be limited through management methods (e.g., by configuring limits), communication performance will still degrade if available radio resources cannot meet all network traffic demands. There are various types of mobile robots and manipulators on the market with highly sophisticated sensing capabilities, some carrying multiple 3D cameras, and these will be deployed in factory floors or medical facilities. Traffic from groups of such robot units can create significant resource pressure when sensing data needs to be processed at edge computing nodes for visual SLAM (Simultaneous Localization and Mapping), robot motion control, and robot task control. If wireless network resources are saturated, sensing data (sent to edge computing nodes) and robot motion references (sent to mobile robots) may not arrive at their destinations in time or may be lost. This prevents mobile robots and manipulators from exhibiting responsive behavior, which is crucial in unstructured work environments.

[0008] Given these limitations and technical challenges, solutions that allow such robotic applications to be supported by edge computing resources at a reasonable capital cost and operating expenditure would be highly valuable. Summary of the Invention

[0009] One object of this disclosure is to provide a control network for supporting mobile robots, which is capable of allocating network resources within an access network to prioritize critical mobile robots and / or critical robot tasks in a systematic and equitable manner. Another object is to avoid costly downtime and emergency stops due to insufficient access network resources. Yet another object is to provide reasonable determinism to ensure that sufficient network resources are available before initiating a robot task. A further object is to enable adaptation during path execution or task execution based on changes in Quality of Service (QoS) over time; such adaptation affects the relative priority of robots or robot tasks, as well as path planning or task planning. A further object of this disclosure is to provide a method for execution within such a control network, for example, by one or more nodes in the control network.

[0010] At least some of these objectives are achieved by the invention as defined in the independent claims. The dependent claims relate to preferred embodiments.

[0011] In a first aspect of this disclosure, a control network is provided for supporting one or more mobile robots operable in a facility. The control network includes a queue management system (FMS) and a network management system (NMS) with mutually exclusive permissions. Specifically, the FMS is authorized to perform path planning and path execution for one or more mobile robots and to assign at least one connection priority to a specific robot or robot task. The NMS is authorized to configure and perform network resource allocation in two or more access networks, wherein each of the two or more access networks is operable to provide wireless connectivity in the facility to a predefined group of one or more mobile robots, and / or to provide such connectivity to the mobile robot while it performs a predefined category of robot task. According to the first aspect, the FMS is configured to generate predictive network resource requests based on path planning and share them with the NMS. The NMS is configured to perform network resource allocation within each of the access networks according to at least one connection priority.

[0012] By using one or more connection priorities, the FMS can influence the distribution of currently available network resources performed by the NMS in a configurable manner. Using connection priorities allows for the systematic and stable distribution of network resources. It is worth noting that a temporary decline in overall network performance (e.g., due to congestion) is generally not a necessary reason to update connection priorities. Another advantage is that the FMS does not need to be granted access to the network and / or the NMS; instead, these components controlling the network can be owned by an external party, or their operation can be outsourced.

[0013] In some embodiments, FMS updates connection priorities based on changes in the QoS of the access network over time, which can be monitored through QoS reports provided by NMS.

[0014] In a second aspect of this disclosure, a method is provided in a control network's Functional Management System (FMS). The method includes: assigning at least one connection priority to a specific robot or robot task; performing planning for one or more mobile robots, including path planning and optional task planning; generating a predictive network resource request based on the output of the planning, and sharing the predictive network resource request with the control network management system (NMS); causing the NMS to allocate network resources in the access network to the one or more mobile robots according to at least one connection priority; and executing the output of the planning, which includes path planning and optional task planning.

[0015] In one embodiment of the second aspect of the method, the planning also includes task planning for at least one of the robots.

[0016] In one embodiment, the second aspect of the method further includes sharing a QoS request with the NMS and deciding to execute a robot task only after receiving a positive response from the NMS. Specifically, the method may include updating at least one assigned connection priority upon receiving a negative response from the NMS. Within this embodiment, the QoS request may be associated with a robot task to be executed within a specified time interval. Additionally or alternatively, the QoS request may include the processing priority required by the robot performing the robot task. Additionally or alternatively, the QoS request may include one or more of the following: estimated bandwidth, data throughput, packet loss rate, packet loss burstiness, transmission reliability, latency, and latency variation.

[0017] In one embodiment, the second aspect of the method further includes receiving a current QoS report from the NMS and performing one or more of the following operations based on the current QoS report: a) updating at least one previously assigned connection priority; b) updating a plan, which includes a path plan; and c) sharing an updated QoS request with the NMS, wherein the updated QoS request replaces the previous QoS request.

[0018] In one embodiment of the second aspect of the method, the output of the execution plan includes feeding control data to the mobile robot and receiving sensor data from the mobile robot.

[0019] In a third aspect of this disclosure, a method is provided in an NMS (Network Management System) for controlling a network. The method includes: receiving at least one connection priority assigned to a specific robot or robot task; receiving a predictive network resource request; and allocating network resources in the access network to one or more mobile robots based on the received at least one connection priority.

[0020] In one embodiment, the third aspect of the method further includes configuring the access network based on at least one received connection priority.

[0021] In one embodiment, the third aspect of the method further includes receiving a QoS request and further allocating network resources based on the QoS request. In this embodiment, the QoS request may be associated with a robotic task to be performed within a specified time interval. Additionally or alternatively, the QoS request may include the required processing priority of the robot performing the robotic task. Additionally or alternatively, the QoS request may include one or more of the following: estimated bandwidth, data throughput, packet loss rate, packet loss burstiness, transmission reliability, latency, and latency variation.

[0022] In one embodiment, the third aspect of the method further includes providing a current QoS report.

[0023] In one embodiment, the third aspect of the method further includes providing connectivity to any mobile robot performing a navigation task. That is, the allocation of network resources is based on whether the mobile robot is performing a navigation task, but can also be independent of location, robot type, and other factors. In a variation of this embodiment, the method further includes providing connectivity to any mobile robot performing a workpiece manipulation task, any mobile robot performing a full-body motion task, and / or any mobile robot performing an emergency task.

[0024] The effects and advantages of the methods in the second and third aspects are substantially similar to those of the method in the first aspect. Furthermore, the methods in the second and third aspects can be implemented with technical changes of the same degree as those in the first and second aspects, as described below.

[0025] This disclosure also relates to a computer program containing instructions for causing a computer (or particularly a control network) to perform the methods described above. This computer program can be stored or distributed onto a data carrier. As used herein, "data carrier" can be a transient data carrier, such as modulated electromagnetic waves or light waves, or a non-transient data carrier. Non-transient data carriers include both volatile and non-volatile memory, such as permanent and non-permanent storage media of magnetic, optical, or solid-state types. Still within the scope of "data carrier," such memory can be fixedly mounted or portable.

[0026] In this disclosure, and particularly in the claims, the term "mobile robot" refers to a self-propelled robot with full-body motion capabilities, including automated guided vehicles (AGVs), autonomous mobile robots (AMRs), and mobile manipulators. Conceptually, a mobile manipulator is a mobile robot with one or more robotic arms.

[0027] Generally, unless otherwise expressly defined herein, all terms used in the claims should be interpreted according to their ordinary meaning in the art. All references to “a / an / the element, device, component, part, step, etc.” should be openly interpreted as referring to at least one instance of that element, device, component, part, step, etc., unless otherwise expressly stated. Unless otherwise expressly stated, the steps of any method disclosed herein need not be performed in the strictly disclosed order. Attached Figure Description

[0028] Various aspects and embodiments will now be described by way of example, with reference to the accompanying drawings, wherein: Figure 1 A control network for supporting a mobile robot queue is shown, which includes an FMS and an NMS. Figure 2 A control network with two independent access networks is shown, which are used to provide wireless connectivity for a predefined group of mobile robots in the facility; and Figure 3 It is a sequence diagram showing the information exchanged between the FMS, NMS, access network, and mobile robot during operation. Detailed Implementation

[0029] Various aspects of this disclosure will now be described more fully with reference to the accompanying drawings, in which certain embodiments of the invention are illustrated. However, these aspects may be embodied in many different forms and should not be construed as limiting; rather, these embodiments are provided by way of example to make this disclosure comprehensive and complete and to fully convey the scope of all aspects of the invention to those skilled in the art. Throughout the specification, the same numerals refer to the same elements.

[0030] Figure 1This is a block diagram of a control network 100 for supporting multiple mobile robots (MRs) 130 operating in a facility. The facility may be a building, factory, plant, warehouse, (partially outdoor) industrial environment, mine, etc. In some embodiments, a public road network is not considered a "facility" in this sense. The control network 100 includes a queue management system (FMS) 110 and a network management system (NMS) 120. The mobile robots 130 reportedly include self-propelled robots with full-body motion capabilities, including automated guided vehicles (AGVs), autonomous mobile robots (AMRs), and mobile manipulators (conceptually, these are mobile robots with one or more robotic arms). The mobile robots 130 can be configured, in particular, as industrial robots or other multi-functional robots, such as medical robots. They can move in two dimensions on a surface or in three dimensions in air or liquid.

[0031] The primary responsibility of NMS 120 is to configure multiple access networks 125 and perform resource allocation within them. NMS 120 may have dedicated permissions to configure access networks 125 at runtime. Each access network 125 includes a wireless access point (WAP) installed in the facility to provide wireless connectivity for the mobile robot 130. This connectivity is provided via an air interface, each extending from one of the WAPs to a wireless field device (WFD) 135 within the mobile robot 130. The WFD 135 may constitute a (mobile) station for Wi-Fi™ (IEEE 802.11 series), a field device for WIA-FA, or a user equipment (UE) suitable for cellular networks (e.g., 5G networks, such as 3GPP NG-RAN). Accordingly, the WAP may be a Wi-Fi™ access point, a WIA-FA access device, or a base station (NB, eNB, gNB) for a cellular network. The mobile robot 130 may connect to different WAPs as it moves between different areas of the facility, which can be achieved through handover techniques known per se. Additionally, access network 125 can be used to transmit instructions from FMS 110 to mobile robot 130, including planned motion paths, robot tasks (practical tasks), and robot motion references. Example robot tasks include navigation, workpiece manipulation, and full-body motion. Multiple robot tasks can be presented as a single robot task. The transmission takes the form of an ordered set of robotic tasks. These tasks do not overlap or (partially) intersect in time: .

[0032] exist Figure 1The diagram shows an example of the internal structure of the NMS 120. This structure (similar to the structure of the FMS 110 described below) should be understood as a functional structure. More precisely, it can correspond to the arrangement of physical components or the arrangement of executable software code sections. Figure 1 The connectors in the diagram represent typical information flows that occur when the NMS 120 operates in accordance with this disclosure. The layout of these flows does not preclude information from traveling along other paths within the internal structure, for example, when multiple components are connected to a public network with a star or mesh topology.

[0033] In NMS 120, network monitor 122 collects state information from WAP in access network 125 and WFD 135 in MR 130 within the facility and stores it in network database 123. In many cases, even if WAP and WFD 135 are deployed and owned by the same entity as MR 130 and the facility, FMS 110 cannot directly access the underlying state information, which is the primary basis for predicting network QoS. Restricted access to this underlying state information is particularly common in newer wireless technologies, including 5G, WIA-FA, and Wi-Fi 6.

[0034] Furthermore, in NMS 120, network configurator 121 configures WAP and WFD 135 according to the resource allocation plan. Network configurator 121 can access the resource allocation plan in network database 123.

[0035] The network resource allocator 127 in NMS 120 generates a resource allocation plan based on collected status information related to the wireless network retrieved from network database 123. Additionally, a resource allocation plan can be generated based on a facility map provided by map database 124. Furthermore, a resource allocation plan can be generated based on predictive network resource requests received from FMS 110, as described in more detail below. The network resource allocator 127 stores the generated resource allocation plan in network database 123.

[0036] The NMS 120 also includes a network QoS predictor 126. The QoS predictor 126 is configured to predict achievable network QoS based on resource allocation plans and optionally on facility-related map information.

[0037] Moving to FMS 110, this system, controlling network 100, is responsible for path planning and execution for MR 130. FMS 110 can also handle task planning and execution, assigning practical tasks (production, processing, material handling, transportation, etc.) to one or more of the mobile robots 130. These tasks may be part of a higher-level task or project to be performed by FMS 110. Path planning may support the assigned tasks or may be non-productive, such as those related to the maintenance or parking of MR 130. Path execution may include providing movement commands (motion references) wirelessly to the mobile robots 130 via access network 125. FMS 110 may have dedicated permissions to perform path planning and execution during runtime. FMS 110 or parts thereof may execute Robot Operating System 1 (ROS1) or Robot Operating System 2 (ROS2).

[0038] A queue monitor 112 is provided within the FMS 110. The queue monitor 112 collects status information from the MR 130 and stores the status information in the queue database 113.

[0039] Task planner 118 generates tasks to be performed by MR 130 based on higher-level inputs (e.g., production plans from manufacturing execution systems (not shown)).

[0040] Path planner 117 generates a path for MR 130 and stores the path in queue database 113. Path planner 117 can be configured to generate the path based on the collected status information of MR 130 retrieved from queue database 113 and the tasks assigned to MR 130 by task planner 118. Path planning can also be based on a facility map, which can be obtained from map database 114. Alternatively or additionally, path planning can also be based on the achievable (predicted) QoS of the wireless network indicated by QoS predictor 126 in NMS 120.

[0041] Furthermore, in FMS 110, path executor 111 controls MR 130 to implement the path generated by path planner 117 from queue database 113 (as seen above), while taking into account the state of mobile robot 130 from queue database 113.

[0042] A network demand forecaster 116 is also provided, which is operable to predict the quantity, type, and / or location of required network resources. The network demand forecaster 116 can make this prediction based on the MR routes and facility maps determined by the route planner 117 (the latter can be obtained from the map database 114).

[0043] Relative to at least one of the mobile robots 130, the FMS 110 can be described as an edge computing resource. That is, the processing circuitry in the FMS 110 is positioned relative to the topology of the access network 125 so that the mobile robot 130 can enjoy reasonable QoS under normal circumstances; for example, the connection between the mobile robot 130 and the FMS 110 typically meets minimum throughput, maximum permissible latency, or similar requirements. Under normal circumstances, these QoS requirements can be met by appropriately positioning the FMS 110 relative to the mobile robot 130 and / or configuring parameters related to routing, scheduling, resource allocation, and service prioritization in the access network 125.

[0044] Communication between FMS 110 and NMS 120 can be achieved using traditional unicast, multicast, or broadcast messages. Alternatively, such as... Figure 1 As illustrated, communication can be supported by publish-subscribe (or PubSub) services 140 and 150. Publish-subscribe services 140 and 150 are configured to facilitate the sharing of assigned connection priorities, predictive network resource requests and responses, QoS requests and / or QoS reports between FMS 110 and NMS 120.

[0045] The publish-subscribe services 140 and 150 are characterized by asynchronous message exchange, allowing the receiver to receive messages independently of (and possibly later than) the time the sender publishes the message. Furthermore, compared to traditional unicast or broadcast transmissions initiated by the sender and message polling by the receiver instance, the publish-subscribe services 140 and 150 feature a greater degree of decoupling between the sender and receiver. In practice, the sender (or publisher) can publish messages to a storage device that belongs neither to itself nor to the receiver (or subscriber), although this storage device is provided by an intermediary (e.g., middleware for receiving and sending unicast or multicast messages). Implementations of this invention may conform to the Data Distribution Service (DDS) standard or the Open Platform Communications Unified Architecture (OPC UA) PubSub protocol, or alternatively, if a cellular connectivity framework is applied, to the 3GPP 5G / NR Network Open Functions (NEF) Northbound Interface specification. Alternatively or additionally, some implementations of the present invention may comply with the Service Enablement Architecture Layer (SEAL) of 3GPP Release 16 and later, particularly the device-centric capabilities related to network resource management, location information, etc.

[0046] As described in the applicant's WO2022214193 cited above, publish-subscribe services 140 and 150 can be configured as a network demand topic 140 and a network QoS topic 150. In a specific implementation, FMS 110 instantiates publisher 141 in network demand topic 140 and subscriber 152 in network QoS topic 150. Similarly, NMS 120 instantiates publisher 151 in network QoS topic 150 and subscriber 142 in network demand topic 140. With this setup, the path planner 117 in FMS 110 can access information related to achievable QoS through the subscribed network QoS topic 150, or more precisely, through subscriber 152. The network demand forecaster 116 predicts the required network resources and publishes the corresponding information, making this information available to NMS 120 through network demand topic 140. Furthermore, the network resource allocator 127 in NMS 120 generates a resource allocation plan based on predictive resource requests from FMS 110, which are received through subscribed network demand topics 140. The network QoS forecaster 126 publishes its output using network QoS topics 150 for use by FMS 110.

[0047] In this disclosure, the inventors propose using the type and capabilities of mobile robots 130 (including mobile robot manipulators) as a guide to plan the operation of the entire wireless infrastructure in a factory floor, medical facility, or warehouse. For example, a dedicated access network can be set up to serve only the mobile robot manipulators, while another access network is used to provide connectivity for autonomous mobile robots (AMRs). This separation reduces the likelihood of communication performance degradation due to too many wireless clients serving the same access network. To keep the overall infrastructure cost reasonable, multiple such wireless networks can be established on the same device. In various embodiments, different service sets associated with specific radio frequency bands and channels can be defined in WLAN / Wi-Fi access points, while different logical / virtual networks or network slices can be set up in cellular-based systems (e.g., 4G / LTE or 5G / NR).

[0048] Figure 2 A control network 100 with two independent access networks 125-1 and 125-2 is shown, which are used to provide wireless connectivity for different groups of mobile robots 130 in the facility. It can be seen that two mobile robots 130-1 and 130-3 are served by the first access network 125-1, while two other mobile robots 130-2 and 130-4 are served by the second access network 125-2. Figure 2The diagram illustrates that the allocation of access network 125 to mobile robot 130 is not necessarily based on proximity (e.g., a rule, i.e., the access network 125 with the least signal attenuation should be selected), although in some embodiments, this association may be subject to a requirement that the signal power must exceed a lower threshold. This allocation is controlled by NMS 120, which accordingly instructs the network nodes of the corresponding access networks 125-1, 125-2. This instruction may be provided based on contextual information, such as the type and capabilities of mobile robot 130 and / or the category and urgency of the robot task assigned to mobile robot 130 and / or the planned robot trajectory.

[0049] More precisely, the category and urgency of the robot task, along with the planned robot trajectory, can be used to differentiate service processing between wireless clients within the access network. In fact, not all mobile robots 130 require the same Quality of Service (QoS) communication, especially in terms of latency and bandwidth, nor do they require the same QoS at all stages of operation. For example, mobile robot 130 may perform navigation tasks, manipulation tasks, and full-body motion tasks, or combinations thereof, at different times. Navigation tasks typically require only lower network bandwidth, while manipulation and full-body motion tasks can tolerate higher latency. QoS requirements may be related to (proportional to) the frequency with which motion references are sent to mobile robot 130. Therefore, for each mobile robot, the required communication QoS can be adjusted based on the assigned robot task it is performing and the priority of that task. Such factors (contextual information) can be utilized when reserving network bandwidth and specifying service processing rules.

[0050] From a high-level perspective, the following approach can be followed: Before FMS 110 assigns a robot task to mobile robot 130, FMS 110 generates a QoS request and shares it with NMS 120. Based on the type and capabilities of mobile robot 130 and / or the level and urgency of the robot task, FMS 110 determines the wireless network processing priority required for the QoS request and the estimated bandwidth, etc. Next, FMS 110 will query NMS 120 for the current operating status of the relevant access networks (multiple) 125 and whether the QoS request can be met. If the access network conditions are found to be unsupported for the normal execution of robot tasks, the FMS 110 may choose to postpone the assignment of robot tasks until the access network is restored or the congestion level is reduced. Otherwise, FMS 110 calculates a path (or trajectory) for mobile robot 130. NMS 120 reserves rated bandwidth in all access points of access network 125 accessed along the calculated path and configures their service processing parameters. Following this stage, FMS 110 begins streaming the necessary commands to mobile robot 130 in real time. In doing so, FMS 110 can also receive sensor data streams from mobile robot 130, thereby handling obstacles or other unforeseen events in the mobile robot's path.

[0051] As a practical example, services performed by a mobile robot during navigation tasks in a non-chaotic environment can be assigned lower priority without causing harm, allowing other mobile robots to benefit from higher communication QoS during phases with higher wireless communication performance requirements. Typically, robot tasks requiring significant communication bandwidth are full-body movements with visual servoing. In a particular embodiment, WLAN / Wi-Fi access point parameters can be configured, allowing service flows to be categorized and specific service processing applied to each wireless client (and thus, each mobile robot 130). Service flows can be categorized based on source and destination IP addresses, the Transmission Control Protocol (TCP) or User Datagram Protocol (UDP) transport protocol port in use, Differential Service Code Point (DSCP) values ​​in data packets, etc. Different QoS levels can then be defined and assigned to the relevant service flows.

[0052] Now for reference Figure 3 The sequence diagram in the diagram describes a method for operating a control network 100. Each step of this method will be executed in an FMS 110, an NMS 120, access networks 125-1, 125-2 (NW1, NW2), or a combination of these network nodes. Of course, the execution of this method will also affect the mobile robot 130, for example, affecting the control data to be executed by the actuators in the mobile robot 130, sensor data read from the sensors in the mobile robot 130, and / or configuration data to be loaded into the memory or processor of the mobile robot 130. It should be emphasized that the exact order of these steps may vary. Figure 3 The examples depicted are different, and one or more steps may overlap in time.

[0053] As explained above, the main responsibilities of FMS 110 may include: - Track mobile robot 130, its capabilities, robot tasks to be assigned, and connection priorities; - Manage and adjust the relative connection priorities of mobile robots 130 and the robot tasks that can be assigned to them; - Plan the robot path (or trajectory) and send robot motion references to execute the path; - Optionally, based on the status report about the access network 125 received by FMS 110 from NMS 120, a QoS request related to the robot task can be calculated; and the QoS request can be shared with NMS 120. The main responsibilities of NMS 120 may include: - Receive QoS requests from FMS 110 and configure the required QoS levels and service processing rules in the infrastructure (including connection points) of access network 125 as much as possible; and - Report the current status of access network 125 to FMS 110.

[0054] Within this method, such as Figure 3 As shown, FMS 110 can assign at least one connection priority 310 to a mobile robot 130 or robot task. Connection priority 310 can be assigned collectively, for example, by referring to the type of robot group or other public identifier. Similarly, connection priority 310 can be assigned to a robot task group. Furthermore, for a single mobile robot 130, connection priority 310 can have multiple independent values; that is, for each access network 125 used by a particular mobile robot 130, connection priority 310 has one value. Depending on any of these choices, the assigned at least one connection priority 310 will be shared with NMS 120.

[0055] Connection priority 310 refers to the order in which services are provided to different mobile robots or different robot tasks (i.e., the allocation of requested QoS). This order can be expressed as a list where natural numbers are assigned to different mobile robots or different robot tasks, meaning that, in terms of network resource allocation, robots / tasks with lower assignment numbers take precedence over robots / tasks with higher assignment numbers. This form can simplify subsequent conversion to standardized network priority parameters such as DiffServ code points (DSCP), WLAN / Wi-Fi user preferences, and 5G QoS identifiers (QCI) if needed.

[0056] NMS 120 is responsible for configuring each of the access networks 125. Such configuration 312 may involve defining the internal operations of the access network nodes and / or their operational parameters (settings) relative to the behavior of connected wireless clients. Different configurations can be applied to different access networks 125. Network configuration 312 may include association rules, such as which predefined mobile robot group 130 and / or which predefined robot task category will be served by which access network 125. The configuration of access network 125 may depend on the connection priority 310 assigned by FMS 110. Configuration 312 can be delegated to the network configurator 121 within NMS 120.

[0057] FMS 110 also performs various types of planning. These plans may include path planning 314, but in some embodiments of the method, task planning 316 is also included. As described above, path planning 314 may include the use of navigation algorithms, such as visual SLAM (Simultaneous Localization and Mapping), and its output may be a path or trajectory to be followed by at least one mobile robot 130. Path planning 314 may be the responsibility of the path planner 117 of FMS 110. Similarly, task planning 316 relates to robot tasks (e.g., workpiece manipulation, transportation) that will be performed by a robotic arm or robot manipulator carried by at least one mobile robot 130, or, in the case of transportation tasks, by the full-body movement of the mobile robot 130. Task planning 316 may be the responsibility of the task planner 118 of FMS 110, or it may be assigned to another entity in the control network 100, or even to an external entity. Optionally, the output of the planning may be temporarily stored in FMS 110 (e.g., in a queue database 113) to allow for later execution.

[0058] Before executing a planned path or a planned robotic task, in some embodiments, FMS 110 may query NMS 120 whether the access network 125 can provide a sufficient QoS level of connectivity during execution. To this end, FMS 110 may generate a QoS request 318 and share it with NMS 120, expecting NMS 120 to respond by sharing a response (or QoS prediction) 320 with FMS 110. The network QoS predictor 126 in NMS 120 may be configured to generate the response 320.

[0059] QoS request 318 depends on or is based on the output of planning phases 316 and 318. Part of its content can be prepared by the network demand forecaster 116 described above. Specifically, QoS request 318 may relate to robot tasks that one or more mobile robots 130 will perform within a specified time interval. Optionally, QoS request 318 may also specify the approximate location where the robot tasks will be performed, such as a part of a factory. It is understood that QoS requirements or priorities can be assigned based on previously assigned connection priorities 310.

[0060] Another option for QoS request 318 is to specify the processing priority required by the mobile robots(s)130 performing the robot task. The processing priority can be a relative priority, indicating the extent to which this(s) mobile robot(s)130 should take precedence over other mobile robots(s)130 when the NMS 120 must decide on network resource contention requests from different mobile robots(s). During operation, the absolute QoS metric corresponding to this processing priority typically depends on the current operating conditions of the access network 125. Nevertheless, the inventors have found that ensuring a certain (high) processing priority can reasonably ensure the successful execution of a particular robot task. This is especially true in a control network 100 that operates stably and experiences minimal daily fluctuations. Furthermore, QoS request 318 can optionally specify one or more absolute QoS metrics (rather than the (relative) processing priority), such as estimated bandwidth, data throughput, packet loss rate, packet loss burstiness, transmission reliability, latency, latency variation, maximum permissible latency, and minimum required transmission reliability.

[0061] FMS 110 may decide whether to execute the robot task based on the content of response 320. If response 320 is affirmative (e.g., the network resources for the requested QoS are indeed available), FMS 110 will typically proceed to the execution phase, including path execution 328 and optional task execution 330. Path execution 328 may be performed by or with the assistance of path executor 111 in FMS 110. If response 320 is negative, in some embodiments, FMS 110 may decide to retry later using the same planned path or planned task. In other embodiments, FMS 110 may be configured to replan the path and / or task to provide it with more moderate QoS requirements; then, FMS 110 resubmits the QoS request 318 to NMS 120, and an increased likelihood of receiving a positive response 320 can be expected. The replanning option may require greater computation, but this is considered reasonable for urgent tasks. In some other embodiments, FMS 110 may take different actions in response to a non-fully affirmative QoS response 320 from NMS 120, such as waiting and retrying if the QoS deficiency is relatively large, and rescheduling if the QoS deficiency is relatively small. A threshold can be defined for this purpose. A similar response can be configured if NMS 120 fails to send response 320 within a predefined time.

[0062] Furthermore, the content of response 320 (especially negative or partially negative content) in some embodiments can cause FMS 110 to update at least one connection priority assigned by 334. This is especially true if QoS inadequacy is found to persist (e.g., after several retries by FMS 110), indicating that one of the access networks 125 has insufficient resources and that these resources need to be distributed among the mobile robots 130 in a different manner. In particular, more urgent or important tasks typically take precedence over regular tasks.

[0063] When FMS 110 is about to execute a planned path or a planned robot task, after making an optional QoS query (messages 318, 320) with NMS 120, FMS 110 submits a predictive network resource request 322 to NMS 120. This predictive network resource request 322 is based on the output of the planning phase, such as information obtained from the planned path or planned task. The predictive network resource request 322 may include a valid time interval or an expiration time, allowing NMS 120 to release the requested network resources and reallocate them to other mobile robots 130 or other robot tasks. Generally, the structure of the predictive network resource request 322 can be similar to that of QoS request 318 in terms of scope (robot task, time interval, location) and substantive QoS requirements. The predictive network resource request 322 may optionally reference information from a previous QoS request 318 to avoid information duplication.

[0064] In response to the predictive network resource request 322, the NMS 120 anticipates performing a network resource allocation 324 associated with one or more access networks 125. The resource allocation 324 can be performed by configuring the operating parameters of the access networks 125. Within the NMS 120, the network resource allocator 127 can be responsible for performing the resource allocation 324. Besides the predictive network resource request 322 itself and other possible factors, the resource allocation 324 may also depend on at least one assigned connection priority 310 and / or the content of a previous QoS request 318. This dependency is determined by… Figure 3 The curved dashed arrow indicates this. In this sense, FMS 110 enables NMS 120 to allocate network resources in access network 125 to one or more mobile robots 130 based on at least one connection priority 310.

[0065] Multiple access networks 125 will sequentially provide connectivity 326 to mobile robot 130. The mobile robot 130 provided with connectivity may be performing a navigation task, a workpiece manipulation task, and / or a full-body motion task. NMS 120 may also be configured to provide dedicated connectivity via access networks 125 to mobile robots 130 performing emergency tasks (i.e., tasks with a higher level of urgency than normal tasks). The connectivity 326 provided by access networks 125 may persist until further notice; that is, it may proceed in parallel with subsequent message exchanges between FMS 110, NMS 120, and / or mobile robot 130.

[0066] At this point, FMS 110 can enter the execution phase, where the planned outputs are executed, including path execution 328 and / or task execution 330. The execution phase may include feeding control data to the mobile robot 130 and receiving sensor data from the mobile robot 130. The control data can be determined using an open-loop method (e.g., script execution) or a closed-loop method, in which sensor data is input into a predefined control law that generates the control data in a manner known per se. As mentioned above, sensor data streams can consume significant bandwidth, for example, if the mobile robot 130 operates one or more cameras. The control data itself can be expressed as a robot motion reference transmitted at a predefined frequency. For example, a motion reference frequency of 20 Hz might be sufficient for path execution, while 250 Hz might be suitable for performing industrial assembly tasks.

[0067] Understandably, in some embodiments, network resource allocation 324 may be updated during the execution phase, for example, as mobile robot 130 switches from a first task with lower QoS requirements to a second task with higher QoS requirements, or vice versa. This task switching may occur between FMS 110 and NMS 120, or it may be implicit in a pre-agreed planned path and / or a planned task schedule, or NMS 120 may use sensors (network detectors) or other indirect information to monitor the progress of the execution phase.

[0068] NMS 120 can provide FMS 110 with feedback on the operational status of access network 125 during the execution phase. This allows FMS 110 to interrupt or modify path execution 328 or task execution 330 in the event of unexpected deterioration of relevant conditions in access network 125. Otherwise, the feedback confirms to FMS 110 that QoS meets the forecast, reassuring FMS 110 that the execution phase can continue as planned. Specifically, NMS 120 can provide feedback by making QoS report 332 available to FMS 110. QoS report 332 can indicate the current value of one or more QoS metrics, such as bandwidth, data throughput, packet loss rate, packet loss burstiness, transmission reliability, latency, and latency variation. These values ​​can be measurements, including radio measurements or network pings, or indirect estimates of QoS metrics. In particular, network monitor 122 in NMS 120 can provide the feedback contained in QoS report 332. QoS report 332 can be restricted to a specified area of ​​the facility and / or have a specified validity period starting from the current time.

[0069] FMS 110 can be configured to take no action when it receives a QoS report 332 with the expected content; this can be understood as path execution 328 and / or task execution 330 continuing as planned. However, if QoS report 332 indicates that the QoS has significantly decreased compared to the expected level (e.g., as indicated in FMS 110's response 320 to a previous QoS request 318), FMS 110 can react, for example, by updating 334 at least one previously assigned connection priority. Furthermore, FMS 110 can be configured to react by updating 336 the planning (including path planning 314), where the results of the new planning will be executed for the remainder of the execution phase. Additionally, FMS 110 can also react by sharing an updated QoS request 338 with NMS 120; if network resource allocation 324 is being executed based on a previous QoS request 318, the updated QoS request 338 will replace that execution, causing a quantitative or qualitative change in network resource allocation 324.

[0070] After any of the updates mentioned above, 334, 336, or 338 (although this may be in...) Figure 3 (Not explicitly stated in the text) The execution phase can continue with new settings. Furthermore, after updates 334, 336, and 338, new cycles of path planning 314 and / or task planning 316 can be initiated, and the outputs of these cycles can be executed similarly to the workflow described above. FMS 110 is typically configured to continue execution phases uninterrupted as it receives and analyzes QoS reports 332; this helps avoid costly downtime. If QoS report 332 indicates a more serious persistent QoS deficiency, FMS 110 can temporarily pause the execution phase.

[0071] The foregoing has primarily described various aspects of this disclosure with reference to several embodiments. However, those skilled in the art will readily recognize that other embodiments besides those disclosed above are also possible within the scope of this invention (as defined in the appended claims).

Claims

1. A control network (100) for supporting one or more mobile robots (130) operable in a facility, the control network comprising: A queue management system (FMS) (110) is authorized to perform path planning and path execution for the one or more mobile robots and is authorized to assign at least one connection priority to a specific robot or robot task; as well as A network management system (NMS) (120) is authorized to configure and perform network resource allocation in two or more access networks (125), each of which is operable to provide wireless connectivity in the facility to a predefined group of the one or more mobile robots, and / or to provide such connectivity to the mobile robots while they are performing predefined categories of robot tasks. The permissions of the FMS and the NMS are mutually exclusive. The FMS is configured to generate predictive network resource requests (322) based on the path planning and to share the predictive network resource requests (322) with the NMS. The NMS is configured to perform network resource allocation (324) in each of the access networks according to the at least one connection priority.

2. The control network (100) according to claim 1, wherein: The FMS (110) is configured to generate a Quality of Service (QoS) request (318) related to the robot task and share the QoS request with the NMS (120); The NMS is configured to predict (320) whether the QoS request can be satisfied and indicate the prediction to the FMS; as well as The FMS is also configured to decide to execute the robot task only upon receiving an affirmative indication from the NMS.

3. The control network (100) according to claim 2, wherein the QoS request (318) is related to a robot task to be performed within a specified time interval.

4. The control network (100) according to claim 2 or 3, wherein the QoS request (318) includes the required processing priority of the robot (130) performing the robot task.

5. The control network (100) according to any one of claims 2 to 4, wherein the QoS request (318) includes one or more of the following: estimated bandwidth, data throughput, packet loss rate, packet loss burstiness, transmission reliability, delay, delay variation, and maximum permissible delay with minimum required transmission reliability.

6. The control network (100) according to any one of claims 2 to 5, wherein the NMS (120) is further configured to allocate network resources according to the QoS request (318) if the FMS decides to perform the robot task.

7. The control network (100) according to any one of the preceding claims, wherein: The NMS (120) is configured to generate a current QoS report (332) and share the current QoS report (332) with the FMS (110); as well as The FMS (110) is configured to update at least one previously assigned connection priority (334) based on the current QoS report (332).

8. The control network (100) according to claim 7, wherein the FMS (110) is further configured to update the path planning (336) based on the current QoS report (332).

9. The control network (100) according to claim 7 or 8, wherein the FMS (110) is further configured to update the previously generated QoS request (338) based on the current QoS report (332).

10. The control network (100) according to any one of the preceding claims, wherein the FMS (110) is configured for path execution (328), the path execution (328) including feeding control data to the mobile robot and receiving sensor data from the mobile robot (130).

11. The control network (100) according to any one of the preceding claims, wherein the NMS (120) is configured to provide connectivity to one of the following groups: any mobile robot (130) performing a navigation task; any mobile robot (130) performing a workpiece manipulation task; any mobile robot (130) performing a full-body motion task; any mobile robot (130) performing an emergency task.

12. The control network (100) according to any one of the preceding claims, wherein the NMS (120) is configured to: dedicate at least one of the access networks (125) to a predefined group of the one or more mobile robots (130), and / or dedicate at least one of the access networks to mobile robots performing robot tasks of a predefined category.

13. A method for supporting one or more mobile robots (130) operable in a facility equipped with two or more access networks (125), wherein the method is implemented in a queue management system (FMS) (110) in a control network (100), and the method comprises: Assign at least one connection priority to a specific robot or robot task (310). A planning process is performed for the one or more mobile robots, the planning including path planning (314). Based on the output of the plan, a predictive network resource request (322) is generated and shared with the network management system NMS (120); The NMS allocates (324) network resources in the access network to the one or more mobile robots according to the at least one connection priority; as well as The output of the plan (328) is executed, the plan including the path plan.

14. A method for supporting one or more mobile robots (130) operable in a facility equipped with two or more access networks (125), wherein the method is implemented in a network management system (NMS) (120) in a control network (100), and the method comprises: Receive at least one connection priority assigned to a specific robot or robot task (310). Receive predictive network resource requests (322); as well as Based on the received at least one connection priority, network resources in the access network are allocated (324) to the one or more mobile robots.

15. A computer program comprising instructions that cause the control network (100) according to claim 1 to perform the method according to claim 13 or 14.

Images