WO2022060266A1 - Nœud de réseau et procédé de gestion de l'inspection de structure - Google Patents

Nœud de réseau et procédé de gestion de l'inspection de structure Download PDF

Info

Publication number
WO2022060266A1
WO2022060266A1 PCT/SE2020/050875 SE2020050875W WO2022060266A1 WO 2022060266 A1 WO2022060266 A1 WO 2022060266A1 SE 2020050875 W SE2020050875 W SE 2020050875W WO 2022060266 A1 WO2022060266 A1 WO 2022060266A1
Authority
WO
WIPO (PCT)
Prior art keywords
network node
model
indications
network
parameters
Prior art date
Application number
PCT/SE2020/050875
Other languages
English (en)
Inventor
Swarup Kumar Mohalik
Ajay Kattepur
Aneta VULGARAKIS FELJAN
Marios DAOUTIS
Original Assignee
Telefonaktiebolaget Lm Ericsson (Publ)
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Telefonaktiebolaget Lm Ericsson (Publ) filed Critical Telefonaktiebolaget Lm Ericsson (Publ)
Priority to PCT/SE2020/050875 priority Critical patent/WO2022060266A1/fr
Publication of WO2022060266A1 publication Critical patent/WO2022060266A1/fr

Links

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/10—Simultaneous control of position or course in three dimensions
    • G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/70—Services for machine-to-machine communication [M2M] or machine type communication [MTC]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G06N5/022—Knowledge engineering; Knowledge acquisition
    • G06N5/025—Extracting rules from data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/38—Services specially adapted for particular environments, situations or purposes for collecting sensor information

Definitions

  • Embodiments herein relate to a network node and a method performed therein. Furthermore, a computer program product and a computer-readable storage medium are also provided herein. In particular, embodiments herein relate to handling inspection of a structure such as identifying an equipment and/or inspecting the structure such as a communication installation e.g. an antenna site or a power line, for faults in the communications network.
  • a structure such as identifying an equipment and/or inspecting the structure such as a communication installation e.g. an antenna site or a power line, for faults in the communications network.
  • mobile devices also known as wireless communication devices, mobile stations, aerial devices, vehicles, stations (ST A) and/or wireless devices, communicate with one or another or with a server or similar via a Radio access Network (RAN) to one or more core networks (CN).
  • the RAN covers a geographical area which is divided into service areas or cell areas, with each service area or cell area being served by a radio network node such as an access node e.g. a Wi-Fi access point or a radio base station (RBS), which in some radio access technologies (RAT) may also be called, for example, a NodeB, an evolved NodeB (eNodeB) and a gNodeB (gNB).
  • RAT radio access technologies
  • the service area or cell area is a geographical area where radio coverage is provided by the radio network node.
  • the radio network node operates on radio frequencies to communicate over an air interface with the wireless devices within range of the access node.
  • the radio network node communicates over a downlink (DL) to the wireless device and the wireless device communicates over an uplink (UL) to the access node.
  • the radio network node may comprise one or more antennas providing radio coverage over one or more cells.
  • the one or more antennas may be mounted on a radio tower or another structure to enhance the coverage.
  • SUBSTITUTE SHEET seen as more agile, cost-effective, safer and efficient solutions which can help when carrying out remote site inspection tasks. These tasks usually involve moving to the site components of interest, taking measurements and images and getting these artefacts processed for possible mitigating actions.
  • the measurements and images when processed, may dynamically generate more goals, such as objectives for e.g. taking a zoomed-in image, taking signal measurements at adjacent components, checking the microwave line-of-sight for obstructions etc., which must be fulfilled within a mission timeline with available resources.
  • objectives for e.g. taking a zoomed-in image, taking signal measurements at adjacent components, checking the microwave line-of-sight for obstructions etc., which must be fulfilled within a mission timeline with available resources.
  • UAVs and MIDs are yet to reach the levels of efficiency and autonomy required to truly optimize inspection operations, for example due to the limited battery capacity or the limited payloads they can load or carry. This becomes a harder challenge since on-board computational capacity of the devices may be limited whereas machine learning
  • a mobile device e.g. a vehicle/robot/hand-held inspection device, may have constrained memory, power, and compute power which resources may further be drained by running inference e.g. machine learning inference.
  • the mobile device such as a MID might have to resort to taking low-resolution images and/or use low-fidelity models.
  • the MID may use the edge and cloud resources if the models are available or can be downloaded and deployed. This option may also not come for free as it may involve estimating the energy necessary for uploading the images to edge or cloud, deploying the models of required quality, using energy required for processing the proposed image of a certain quality, the end-to-end latency for the processing, and above all, this impacts both the latency and the mobile inspection device battery power.
  • QoS Quality of Service
  • An Ultra Reliable Low Latency Communication (URLLC) slice may provide very good latency and bandwidth but may be costly e.g. expensive or exhaust valuable resources.
  • the above scenario portrays the complexity of decision making for the site inspection task.
  • the decisions may also to be revised dynamically and simultaneously adapting to the environmental conditions, e.g. unexpected obstructions by trees or cables, windy or cloudy conditions which hamper taking steady and good images.
  • the expectation is that in spite of all, a site inspection will result in high quality decisions, low false positives and negatives, while ensuring minimum time, total energy consumed and cost for the inspection task.
  • An object of embodiments herein is, therefore, to improve remote inspection of a structure in an efficient manner.
  • the object is achieved by a method performed by a network node for handling inspection of a structure, which inspection is performed by a mobile device in a communications network.
  • the network node obtains indications of parameters relating to network parameters, machine learning (ML) model parameters and mobile device parameters.
  • the network node determines a task plan for performing inspection of the structure based on the obtained indications to achieve a goal requirement; and sends an indication of the determined task plan to the mobile device.
  • ML machine learning
  • the object is achieved by providing a network node for handling inspection of a structure, which inspection is performed by a mobile device in a communications network.
  • the network node is configured to obtain indications of parameters relating to network parameters, ML model parameters and mobile device parameters.
  • the network node is further configured to determine a task plan for performing inspection of the structure based on the obtained indications to achieve a goal requirement; and to send an indication of the determined task plan to the mobile device.
  • a computer program product comprising instructions, which, when executed on at least one processor, cause the at least one processor to carry out any of the methods above, as performed by the network node. It is additionally provided herein a computer-readable storage medium, having stored thereon a computer program product comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out any of the methods above, as performed by the network node.
  • Embodiments herein thus provide a fault inspection planning function to aid mobile devices to plan and deploy the site inspection task.
  • the network node contains knowledge about available resources, goals and algorithmic trade-offs to allow for e.g. energy efficient task planning. Through the use of efficient edge compute and network slicing, optimal processing location, accuracy vs. energy trade-offs and reduction in false alarms during inspection may be achieved.
  • Embodiments herein incorporate information from the obtained indications such as cell locations, sensing, mobile inspection agent capacity and goal requirements to plan task plans, e.g. inspections, in a time bound manner.
  • ML models here may be learned and updated to meet accuracy requirements in an energy efficient manner.
  • Embodiments herein allow deployment of an end-to-end system that integrates sensing, energy efficient processing, re-planning and accurate fault detection.
  • Fig. 1 is a schematic overview depicting an architecture according to embodiments herein;
  • Fig. 2 is a schematic overview depicting a solution according to embodiments herein;
  • Fig. 3 is a combined signalling scheme and flowchart according to embodiments herein;
  • Fig. 4 shows a block diagram depicting some embodiments herein
  • Fig. 5 is a schematic flowchart depicting a method performed by a network node according to embodiments herein;
  • Fig. 6 is a combined signalling scheme and flowchart according to embodiments herein;
  • Fig. 7 is a block diagram depicting a network node according to embodiments herein.
  • Fig. 1 is a schematic overview depicting a communications network 100 wherein embodiments herein may be implemented.
  • the communications network 100 comprises one or more RANs and one or more CNs.
  • the communications network 100 may use any technology such as 5G New Radio (NR) but may further use a number of other different technologies, such as, Wi-Fi, long term evolution (LTE), LTE-Advanced, wideband code division multiple access (WCDMA), global system for mobile communications/enhanced data rate for GSM evolution (GSM/EDGE), worldwide interoperability for microwave access (WiMAX), or ultra mobile broadband (UMB), just to mention a few possible implementations.
  • NR 5G New Radio
  • WCDMA wideband code division multiple access
  • GSM/EDGE global system for mobile communications/enhanced data rate for GSM evolution
  • WiMAX worldwide interoperability for microwave access
  • UMB ultra mobile broadband
  • the communications network 100 may comprise one or more radio network nodes 12 providing radio coverage over a respective geographical area by means of antennas or similar.
  • the radio network node 12 may serve a user equipment (UE) 10 such as a mobile phone or similar.
  • the geographical area may be referred to as a cell, a service area, beam or a group of beams.
  • the radio network node 12 may be a transmission and reception point e.g. a radio access network node such as a base station, e.g.
  • a radio base station such as a NodeB, an evolved Node B (eNB, eNode B), an NR Node B (gNB), a base transceiver station, a radio remote unit, an Access Point Base Station, a base station router, a transmission arrangement of a radio base station, a stand-alone access point, a Wireless Local Area Network (WLAN) access point, an Access Point Station (AP STA), an access controller, a UE acting as an access point or a peer in a Mobile device to Mobile device (D2D) communication, or any other network unit capable of communicating with a UE within the cell served by the radio network node 12 depending e.g. on the radio access technology and terminology used.
  • eNB evolved Node B
  • gNB NR Node B
  • a base transceiver station a radio remote unit
  • an Access Point Base Station such as a NodeB, an evolved Node B (eNB, eNode B), an NR Node B (gNB), a
  • the communications network 100 may further comprise a network node 11 such as a server or application server for collecting and controlling different tasks in the communications network 100.
  • a network node 11 such as a server or application server for collecting and controlling different tasks in the communications network 100.
  • the communications network 100 may comprise a number of other network nodes 13, e.g. an edge node, a core network node or similar, configured to perform computations or similar in the communications network 100.
  • network nodes 13 e.g. an edge node, a core network node or similar, configured to perform computations or similar in the communications network 100.
  • a mobile device 15 is configured to be used to inspect a structure, such as a radio unit, antenna unit, power line, or a radio tower of the radio network node 12.
  • a structure such as a radio unit, antenna unit, power line, or a radio tower of the radio network node 12.
  • Other structures may be any type of tower or building with e.g. mounted equipment.
  • the mobile device 15 is configured to inspect the structure e.g. record an image of the structure.
  • Embodiments herein relate to a fault inspection planning function to aid mobile devices such as the mobile device 15 to plan and deploy the site inspection task.
  • the network node 11 or a fault inspection agent at the network node 11 is informed, e.g. contains knowledge or information, about available resources, goals and algorithmic trade-offs to allow for energy efficient task planning.
  • Through the use of efficient edge computing and network slicing, optimal processing location, accuracy vs. energy tradeoffs, and reduction in false alarms in inspection may be achieved.
  • Embodiments herein may incorporate information about cell locations, sensing, mobile inspection agent capacity and goal requirements to plan inspections in a time bound manner. ML inference algorithms may be known and updated that are able to meet accuracy requirements in an energy efficient manner. This leads to a balance of accuracy and cell energy consumption in mobile inspection tasks.
  • embodiments herein deploy an end-to-end system that integrates sensing, energy efficient processing, re-planning, and accurate fault detection.
  • Fig. 2 is a schematic overview of a scenario for inspection with mobile devices interacting with a NOC and offloading network nodes.
  • Fig. 2 provides a high-level view of the scenario.
  • Mobile inspection units such as vehicles, humans with inspection devices, inspect e.g. antennas and/or power lines for faults.
  • the mobile inspection units are provided goals and work orders from the NOC station over a control network route.
  • Mobile devices are connected to edge and/or cloud nodes for offloading computations over offload network route, and it should be noted that network slices may be generated and/or utilized to efficiently transmit data between stations.
  • Fig. 3 is a combined flowchart and a signalling scheme displaying some embodiments herein.
  • the network node 11 may obtain indications from e.g. a knowledge base, the mobile device 15 and/or from other network nodes. E.g. the network node 11 receives indications of parameters relating to network parameters, ML model parameters and mobile device parameters.
  • the indications may be values, indices, or any suitable representation, indicating processing capacity locations, cell energy consumption, mobile device capabilities, ML models complexity and requirements, and/or any other similar resources.
  • the network node 11 determines the task plan for the mobile device 15 for performing inspection of the structure based on the obtained indications to achieve a goal requirement, e.g. in a time, cost and/or energy efficient manner.
  • the network node 11 may determine a route plan for the mobile device 15 for performing an inspection of e.g. a power line, with an energy consumption below a threshold and within a time interval.
  • the network node 11 may further take e.g. if network slicing is used or not, into account, since network slicing may mean that time interval is reduced but with a high resource cost.
  • the network node 11 sends an indication of the determined task plan to the mobile device 15.
  • the network node 11 may send route plans, actions plans or similar to the mobile device 15.
  • the network node 11 may furthermore determine which ML model and which node or network node to execute the ML model, e.g. for offloading computations such as image processing or similar. This may be based on mobile device capability, network node capability, network slice capability and/or similar.
  • the network node 11 may then transmit to e.g. the network node 13, the ML model or an indication of the ML model.
  • the ML model may already be present at the network node 13 or may be retrievable from another network node.
  • the network node 11 may then e.g. upon a change in the environment, e.g. occurrence of an obstruction or similar, or a change in a goal e.g. a lower latency requirement resulting in a longer time interval, obtain revised indications from the mobile device 15.
  • the network node 11 may then determine an updated task plan dynamically based on the revised indications.
  • the goals may be revised based on observed and processed data, the decisions may also to be revised dynamically simultaneously adapting to the environmental conditions, e.g. unexpected obstructions by trees or cables, windy or cloudy conditions which hamper taking steady and good images.
  • the expectation is that in spite of all, a site inspection will result in high quality decisions (low false positives and negatives) while ensuring minimum time, total energy consumed and cost for the inspection task.
  • the network node 11 may then transmit the revised task plan or indication of task plan to the mobile device 15 or another network node.
  • the network node 11 may further, if necessary, revise the ML model and/or which node to execute the ML model.
  • Fig. 4 shows a Fault Inspection Planning (FIP) Function interfacing with Network Functions and Machine Learning (ML) models.
  • FIP Fault Inspection Planning
  • FIP function which is a cognitive module containing declarative and procedural knowledge about the current goal, network state, topology, plan and execution state, safety and optimization conditions, etc. in its knowledge base (KB).
  • the cognitive module contains in its core, basic reasoning mechanisms such as forward and backward reasoning.
  • a compiler agent which compiles a planning problem from the available knowledge, triggered by any change in the relevant knowledge. Modeling the procedural knowledge for actions is herein provided.
  • An artificial intelligence (Al) planning which generates a plan of action involving actions such as (1) initiating a network slice (2) modifying a network slice of required QoS (3) movement of vehicle (4) measurement actions for the mobile inspection agents - thus controlling and coordinating not only the network resources but also the physical resources, the MID, and its control.
  • NSF Network Slice Selection Function
  • MIA Mobile Inspection Agent
  • NF Network Functions
  • NWDAF Network Data Analytics Function
  • Explanation agents associated with ML models to give feedback to the KB, specifically when ML model outputs are non-conclusive and cannot drive decision.
  • rules may allow exploration with different quality.
  • Reflective set of rules in the KB of the cognitive module are maintained to improve the generated plan to achieve the best of energy vs accuracy tradeoff, e.g. by learning the performance of the ML models and adding associated constraints in the KB.
  • the data regarding energy taken by ML models may be monitored immediately.
  • the data about the accuracy, false positives and negatives, may be collected from related work ticket resolutions and recorded in the KB using a diagnosis function.
  • Embodiments herein disclose the mobile device 15 acquiring images, which mobile device 15, besides comprising a camera, may also have a radiation receiver, e.g. a wideband radio receiver, that may assess e.g. the level of received radio wave energy at different frequency bands corresponding to the emitting frequencies of radiating antennas.
  • a radiation receiver e.g. a wideband radio receiver
  • the network node obtains indications of parameters relating to network parameters, ML model parameters and mobile device parameters.
  • the indications of parameters may comprise one or more of: quality of images, data sizes of images, estimation of communication time, availability and capability of machine learning models, and performance estimation of machine learning models.
  • the network node 11 may further obtain, e.g. receive from the NOC or manually, requirements including goals and inspection locations for the mobile device 15.
  • the indications may be obtained from the KB.
  • the network node 11 determines the task plan for performing inspection of the structure based on the obtained indications to achieve a goal requirement for e.g. achieving the goal in any one or more out of: an energy efficient manner, a time efficient manner, and/or a cost efficient manner.
  • the network node 11 may determine to execute an ML model at a network node, such as the other network node 13, based on the indications to achieve the goal requirement. E.g. the network node 11 may select an ML model and also the network node to execute the ML model e.g. select a Fog node i.e. a network node located at the edge of the communications network to achieve e.g. an energy efficient process. The network node 11 may determine which network node is able to run the ML model in an energy efficient manner, while returning the results within the time bounds.
  • the network node 11 may determine the task plan and execution of the ML model, by performing trade-offs between various types of sensing and accuracy of ML models that are considered to determine task plan and execution of the ML model.
  • the network node 11 may determine which network node is able to run the ML model by further taking into account network slice parameters and energy required to transfer data and ML model to the network node.
  • Action 504. The network node 11 , furthermore, sends an indication of the determined task plan to the mobile device 15.
  • the network node 11 may send the task plan as such and/or an index indicating a behaviour at the mobile device 15.
  • the network node 11 may further send to the network node, i.e. the selected network node, an indication indicating the ML model, e.g. the ML model as such, part of the ML model such as parameter values, and/or an index indicating an ML model at the network node to execute.
  • an indication indicating the ML model e.g. the ML model as such, part of the ML model such as parameter values, and/or an index indicating an ML model at the network node to execute.
  • the network node 11 may further maintain a log of ML models that achieve a goal requirement with a recorded energy consumption.
  • the network node 11 may provide feedback to the knowledge base when outputs from the ML model are non-conclusive. Reflective set of rules in the KB may be updated to improve the determined task plan.
  • the obtained indications may be revised based on observed and processed data, and the determined task plan may be revised dynamically based on the revised indications.
  • the mobile device 15 may first record data such as images with a low granularity and the images may indicate a fault based on output from ML model and the task plan may be revised recommending the mobile device to record data of high granularity and process the recorded data in a core network node with a high processing capacity.
  • Fig. 6 is a combined flowchart and a signalling scheme displaying some embodiments herein.
  • the network node may be a standalone node or a distributed node comprises e.g. a planning agent and a KB.
  • the mobile device 15 is exemplified as a mobile inspection device.
  • the sequence diagram provided in Fig. 6 demonstrates the sequence of interactions that occur between the NOC, FIP, inspection device and the offload compute locations and is illustrating a learning model for energy efficient site inspection.
  • the NOC may transmit goals such as location, time, and/or energy requirements to the network node 11 .
  • the NOC may thus provide high level goals and inspection locations for the inspection agent.
  • the KB may provide indications such as topography, device capacities, compute nodes, to the planning agent.
  • the FIP function service generates a high-level plan based on the requirements.
  • the network node 11 may send information to the mobile inspection device such as indication of location to move, capture, off load location.
  • the KB may transmit or download ML model or protocol to an edge node or cell.
  • the mobile inspection device may sense e.g. recording data such as images with a low granularity.
  • the MID may offload images to the edge node or cell.
  • the edge node or cell may perform execution of the ML model and process the recorded data.
  • the edge node or cell may detect error or fault in the inspected structure. The edge node or cell may then report fault location to the planning agent in the FIP function.
  • the network node may determine revised task plan based on the revised input, that is, location of fault. The network node may then transmit updated information to the MID. Trade-offs between various types of sensing (signal measurement, image capture) and accuracy of models may be considered by the planning agent and MID. Low granularity images are initially preferred, e.g. due to battery and latency considerations. In case further inspection is required, higher granularity sensing may be re-planned at specific locations.
  • the MID may then sense using a higher granularity to confirm the fault.
  • the network node 11 has determined to change the ML model to a more complex model and that should be executed in the cloud (requiring more energy and is more complex).
  • the protocol or the ML model is downloaded to the cloud node.
  • Embodiments herein may consider which of the compute nodes, i.e. network nodes, can run the ML model in an energy efficient manner, while returning the results within the time bounds. On board computation may be battery draining, while the edge compute node may be more energy efficient.
  • the network slice and the energy required to transfer data and ML functions to specific locations may also be included.
  • Varying complexity ML models employ different power characteristics and energy utilization on various hardware processors that may be deployed on mobile inference devices, edge devices or the cloud.
  • a specific object recognition neural network-based model that has a power requirement in the range of approximately 5000mJ may be prohibitive to be repeatedly executed on the on-board processor of a UAV or a MID, where in this case it would not pose any power concern when performing the execution on a network node or a cloud node.
  • the improvements provided by the specific algorithm on various hardware types which can represent energy, latency and computational offloading trade-off, may be taken into account in embodiments herein.
  • the energy efficiency may be traded off with inference accuracy and/or speed for various hardware sets. These considerations may thus be important to consider when planning energy efficient deployments. Such knowledge is incorporated into our planning formalism.
  • the MID offloads images to the cloud node, e.g. the images sensed with high granularity.
  • the cloud node processes images and executes the ML model, and determines whether there is a fault or not at the location based on the output from the ML model.
  • the network node may then raise a fault ticket at the NOC. Since the end goal is to reduce false-positives and/or false-negatives at the NOC, accuracy of the ML models’ inferencing and a log of the models that work best may be maintained. An ML model repository may be maintained that is able to process images within the time bound with desired accuracy.
  • the network node 11 may comprise an arrangement depicted in two embodiments in Fig.7.
  • the network node 11 may comprise a communication interface 700 depicted in Fig.7, configured to communicate e.g. with the communications network 100 also referred to as a cloud network.
  • the communication interface 700 may comprise a wireless receiver (not shown) and a wireless transmitter (not shown) and e.g. one or more antennas.
  • the embodiments herein may be implemented through a processing circuitry 701 configured to perform the methods herein.
  • the processing circuitry may comprise one or more processors.
  • a network node comprising processing circuitry and memory, said memory comprising instructions executable by said processing circuitry whereby said network node 11 is operative to perform the methods herein.
  • the network node 11 may comprise a receiving unit 702, a receiver or a transceiver.
  • the processing circuitry 701 , the network node 11 and/or the receiving unit 702 is configured to obtain the indications of parameters relating to network parameters, ML model parameters and mobile device parameters.
  • the indications of parameters may comprise quality of images, data sizes of images, estimation of communication time, availability and capability of machine learning models and performance estimation of machine learning models.
  • the processing circuitry 701 , the network node 11 and/or the receiving unit 702 may be configured to obtain the indications by obtaining goal requirements including goals and inspection locations for the mobile device from a network operations centre.
  • the processing circuitry 701 , the network node 11 and/or the receiving unit 702 may be configured to obtain the indications from the KB.
  • the network node 11 may comprise a determining unit 703, e.g. a processor.
  • the processing circuitry 701 , the network node 11 and/or the determining unit 703 is configured to determine the task plan for performing inspection of the structure based on the obtained indications to achieve the goal requirement.
  • the processing circuitry 701 , the network node 11 and/or the determining unit 703 may be configured to determine to execute an ML model at a network node based on the indications to achieve the goal requirement.
  • the processing circuitry 701 , the network node 11 and/or the determining unit 703 may be configured to determine the task plan and execution of the ML model, by performing trade-offs between various types of sensing and accuracy of ML models that are considered to determine task plan and execution of the ML model.
  • the processing circuitry 701 , the network node 11 and/or the determining unit 703 may be configured to determine to execute by determining which network node is able to run the ML model in an energy efficient manner, while returning the results within the time bounds.
  • the processing circuitry 701 , the network node 11 and/or the determining unit 703 may be configured to determine which network node is able to run the ML model based on network slice parameters and energy required to transfer data and ML model to the network node.
  • the obtained indications may be revised based on observed and processed data, and the determined task plan may be revised dynamically based on the revised indications.
  • the network node 11 may comprise a transmitting unit 704, e.g. a transmitter or a transceiver.
  • the processing circuitry 701 , the network node 11 and/or the transmitting unit 704 is configured to send the indication of the determined task plan to the mobile device.
  • the processing circuitry 701 , the network node 11 and/or the transmitting unit 704 may be configured to send to the network node the indication indicating the ML model.
  • the processing circuitry 701 , the network node 11 and/or the transmitting unit 704 may be configured to provide feedback to the knowledge base when outputs from the ML model are non-conclusive. Thus, a reflective set of rules in the knowledge base may be updated to improve the determined task plan.
  • the network node 11 may comprise a maintaining unit 705.
  • the processing circuitry 701 , the network node 11 and/or the maintaining unit 705 may be configured to maintain the log of machine learning models that achieve a goal requirement with a recorded energy consumption.
  • the embodiments herein may be implemented through a respective processor or one or more processors, such as a processor of the processing circuitry 701 in the network node 11 depicted in Fig. 7, together with a respective computer program code for performing the functions and actions of the embodiments herein.
  • the program code mentioned above may also be provided as a computer program product, for instance in the form of a data carrier carrying computer program code for performing the embodiments herein when being loaded into the network node 11 .
  • One such carrier may be in the form of a universal serial bus (USB) stick, a disc or similar. It is however feasible with other data carriers such as any memory stick.
  • the computer program code may furthermore be provided as pure program code on a server and downloaded to the network node 11 .
  • the network node 11 may further comprise a memory 770 comprising one or more memory units to store data on.
  • the memory comprises instructions executable by the processor.
  • the memory 770 is arranged to be used to store e.g. ML models, indices, KB information, measurements, photos, location information, meta data, instructions, configurations and applications to perform the methods herein when being executed in the network node 11 .
  • the units in the network node 11 mentioned above may refer to a combination of analogue and digital circuits, and/or one or more processors configured with software and/or firmware, e.g. stored in the network node 11 , that when executed by the respective one or more processors perform the methods described above.
  • processors as well as the other digital hardware, may be included in a single Application-Specific Integrated Circuitry (ASIC), or several processors and various digital hardware may be distributed among several separate components, whether individually packaged or assembled into a system-on-a- chip (SoC).
  • ASIC Application-Specific Integrated Circuitry
  • SoC system-on-a- chip
  • a computer program 790 comprises instructions, which when executed by the respective at least one processor, cause the at least one processor of the network node 11 to perform the actions above.
  • a carrier 780 comprises the computer program 790, wherein the carrier 780 is one of an electronic signal, an optical signal, an electromagnetic signal, a magnetic signal, an electric signal, a radio signal, a microwave signal, or a computer- readable storage medium.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Medical Informatics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

Des modes de réalisation de la présente invention divulguent par exemple un procédé de gestion de l'inspection d'une structure réalisée par un dispositif mobile dans un réseau de communication. Le nœud de réseau obtient des indications de paramètres relatifs à des paramètres de réseau, à l'apprentissage machine, ML, à des paramètres de modèle et à des paramètres de dispositif mobile. Le nœud de réseau détermine un plan de tâches permettant d'effectuer une inspection de la structure sur la base des indications obtenues pour atteindre un objectif ; et envoie une indication du plan de tâches déterminé au dispositif mobile.
PCT/SE2020/050875 2020-09-18 2020-09-18 Nœud de réseau et procédé de gestion de l'inspection de structure WO2022060266A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/SE2020/050875 WO2022060266A1 (fr) 2020-09-18 2020-09-18 Nœud de réseau et procédé de gestion de l'inspection de structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/SE2020/050875 WO2022060266A1 (fr) 2020-09-18 2020-09-18 Nœud de réseau et procédé de gestion de l'inspection de structure

Publications (1)

Publication Number Publication Date
WO2022060266A1 true WO2022060266A1 (fr) 2022-03-24

Family

ID=80776256

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2020/050875 WO2022060266A1 (fr) 2020-09-18 2020-09-18 Nœud de réseau et procédé de gestion de l'inspection de structure

Country Status (1)

Country Link
WO (1) WO2022060266A1 (fr)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9536149B1 (en) * 2016-02-04 2017-01-03 Proxy Technologies, Inc. Electronic assessments, and methods of use and manufacture thereof
US20170206648A1 (en) * 2016-01-20 2017-07-20 Ez3D, Llc System and method for structural inspection and construction estimation using an unmanned aerial vehicle
US20180330238A1 (en) * 2017-05-09 2018-11-15 Neurala, Inc. Systems and methods to enable continual, memory-bounded learning in artificial intelligence and deep learning continuously operating applications across networked compute edges
US20190137995A1 (en) * 2017-11-06 2019-05-09 General Electric Company Systems and method for robotic industrial inspection system
US20190248485A1 (en) * 2018-02-12 2019-08-15 Wipro Limited Method and system for performing inspection and maintenance tasks of three-dimensional structures using drones
US20200004272A1 (en) * 2018-06-28 2020-01-02 Skyyfish, LLC System and method for intelligent aerial inspection
US20200082168A1 (en) * 2018-09-11 2020-03-12 Pointivo, Inc. In data acquistion, processing, and output generation for use in analysis of one or a collection of physical assets of interest

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170206648A1 (en) * 2016-01-20 2017-07-20 Ez3D, Llc System and method for structural inspection and construction estimation using an unmanned aerial vehicle
US9536149B1 (en) * 2016-02-04 2017-01-03 Proxy Technologies, Inc. Electronic assessments, and methods of use and manufacture thereof
US20180330238A1 (en) * 2017-05-09 2018-11-15 Neurala, Inc. Systems and methods to enable continual, memory-bounded learning in artificial intelligence and deep learning continuously operating applications across networked compute edges
US20190137995A1 (en) * 2017-11-06 2019-05-09 General Electric Company Systems and method for robotic industrial inspection system
US20190248485A1 (en) * 2018-02-12 2019-08-15 Wipro Limited Method and system for performing inspection and maintenance tasks of three-dimensional structures using drones
US20200004272A1 (en) * 2018-06-28 2020-01-02 Skyyfish, LLC System and method for intelligent aerial inspection
US20200082168A1 (en) * 2018-09-11 2020-03-12 Pointivo, Inc. In data acquistion, processing, and output generation for use in analysis of one or a collection of physical assets of interest

Similar Documents

Publication Publication Date Title
US11019576B2 (en) Systems and methods for dynamic management of transmit power in a network of moving things
US10659998B2 (en) Systems and methods for managing connectivity between a fixed access point and a hotspot in a network of moving things, for example including autonomous vehicles
KR101782283B1 (ko) 무선 네트워크 조건들에 기반한 계획된 운동의 조정
US20220361011A1 (en) Mismatch detection in digital twin data by comparing to thresholds the reliability and latency constraint communication related values received from sources with geolocation information
Tazibt et al. Wireless sensor network clustering for UAV-based data gathering
EP3632152B1 (fr) Planification du déploiement d'un noeud dans un réseau de communication avec un drone
US9693242B2 (en) Communications links assessment
CN104113524A (zh) 通信系统以及服务器
Fortes et al. Location-based distributed sleeping cell detection and root cause analysis for 5G ultra-dense networks
Amorim et al. Machine-learning identification of airborne UAV-UEs based on LTE radio measurements
US11917422B2 (en) System and method for intelligent throughput performance analysis and issue detection
Chriki et al. Centralized cognitive radio based frequency allocation for UAVs communication
WO2022060266A1 (fr) Nœud de réseau et procédé de gestion de l'inspection de structure
EP4135937A1 (fr) Noeud de réseau, et procédé pour gérer des opérations dans un réseau de communication
CN114980194B (zh) 干扰检测方法、装置及存储介质
US20220095267A1 (en) LINE OF SIGHT (LoS)/NON-LINE OF SIGHT (NLoS) POINT IDENTIFICATION IN WIRELESS COMMUNICATION NETWORKS USING ARTIFICIAL INTELLIGENCE
US20220284822A1 (en) Mobile device and methods for travelling towards a destination using a communication network
JP6394750B2 (ja) 通信システム
WO2023152967A1 (fr) Système de traitement d'informations, dispositif de traitement d'informations, procédé de traitement d'informations et programme
Bennis et al. Short-range and long-range cooperative communication for littoral environment monitoring
Lendinez et al. Enhancing 5G-Enabled Robots Autonomy by Radio-Aware Semantic Maps
US20230343089A1 (en) Computing Device and Method for Handling an Object in Recorded Images
WO2023218543A1 (fr) Terminal, procédé de communication radio, et station de base
WO2024100889A1 (fr) Dispositif de prédiction de qualité de communication, procédé de prédiction de qualité de communication, et programme
CN115980800A (zh) 基于多栖息平台的无人机定位数据通信方法、系统及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20954270

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20954270

Country of ref document: EP

Kind code of ref document: A1