Enhanced position prediction in a communications network

Abstract

The application relates to a method in which a positioning entity configured to predict a position of a mobile entity connected to a cellular network receives from the cellular network, first position related data for a plurality of mobile entities including at least a first mobile entity (200) connected to the cellular network. Furthermore, from the first mobile entity (200), second position related data is received, the second position related data not being related to a communication of the first mobile entity (200) with the cellular network (300). A future position of the first mobile entity is determined based on the first position related data and the second position related data and the future position of the first mobile entity is reported to the cellular network for further processing at the cellular network.

Description

[0001] ENHANCED POSITION PREDICTION IN A COMMUNICATIONS NETWORK

[0002] Technical Field

[0003] The application relates to a method carried out at a positioning entity configured to predict a position of a mobile entity connected to a cellular network or communications network, to the corresponding positioning entity, a computer program comprising program code and a carrier comprising the computer program.

[0004] Background

[0005] The Network Data Analytics Function (NWDAF) is part of 5G core and provides analytics services to other network functions (NFs) and OAM (Operations, Administration and Maintenance) as known from 3 GPP TS 23.288version 16.4.0. The analytics provided by NWDAF includes either statistical information or predictions. Any NWDAF service consumer (i.e., NFs and OAM (Operation Administration and Maintenance) can subscribe to NWDAF analytics service by providing the ID(s) of the demanded analytics. When the output becomes ready, NWDAF notifies the consumer on analytics information.

[0006] The set of analytics provided by NWDAF may comprise: Slice load level, Observed service experience, NF load, Network performance UE (User Equipment) related analytics including UE mobility, UE communication, and Abnormal behavior, User data congestion, QoS sustainability, Dispersion, WLAN (Wireless Local Area Network) performance, Session management, Redundant transmission experience, Data Network performance, PFD (packet flow description) determination, Location Accuracy, End-to-end data volume transfer time and Relative proximity.

[0007] In addition to the network-related analytics and predictions, Rel-18 of 3GPP specifications towards NWDAF proposes to provide advanced analytics about a group of UEs or a specific UE. A NF (e.g., AMF (Access and Mobility Management Function), SMF (Session Management Function)or AF (Application Function)) may consume a UE mobility, UE communication and abnormal behavior analytics provided by NWDAF. To support these analytics and related predictions, NWDAF needs to collect related information from 5G core (5GC) or OAM. For example, in order to generate UE mobility analytics for a service consumer, NWDAF requires data to be collected from AMF. Besides, for UE communication analytics, NWDAF retrieves communication information from SMF, AF, AMF and UPF (User Plane Function). For providing abnormal behavior analytics upon a subscription or request, NWDAF benefits from the data fetched from SMF, AMF, AF and OAM.

[0008] In order to collect corresponding data from NFs, NWDAF can subscribe to the events provided by the NFs. When an event becomes available, NWDAF is informed and notified to collect the new set of data. Before subscribing to the event exposure service provided by an NF, NWDAF first checks the user consent information (i.e., if data is to be collected for a user) by communicating with Unified Data Management (UDM). If user consent is granted, then NWDAF sends a subscription request to the NF. Within this scope, data collection from AF needs to be handled via NEF (Network Exposure Function).

[0009] As stated, NWDAF is capable of retrieving OAM data as well. The procedure of fetching OAM data differs from the NFs. NWDAF needs to request one of the following services provided by OAM for data collection purposes: generic fault supervision or performance assurance management services, Performance Management (PM) services, Fault Management (Fault Management) services.

[0010] Based on this set of input data, NWDAF generates the following information for each type of the following analytics related to the UE positioning:

[0011] • UE mobility analytics o Input: AMF, GMLC (Gateway Mobile Location Centre) o Output: UE location (TAs (tracking areas) or cells) where the UE or UE group may move into or geographical location (longitude and latitude level), Confidence, UE’s direction

[0012] • Location accuracy analytics o Input: AMF, OAM, GMLC o Output: Horizontal and optionally vertical location average accuracy of applicable positioning method, ratio of LOS (Line of Sight)to NLOS ( Non- LOS) measurements of positioning method, ratio of UEs that are indoor / outdoor

[0013] • Proximity analytics o Input: OAM (Speed, Orientation) o Output: Proximity information, Time to collision, accuracy

[0014] • Movement behavior analytics o Input: AMF, GMLC o Output: Average speed, heading direction, total number of users and ratio of moving users

[0015] In addition to the statistical information, NWDAF is capable of generating predictions for each analytic type based on the available data. These predictions are reported along with a confidence value.

[0016] Besides the NWDAF analytics related to UE positioning, 3GPP defines a set of specifications about location services. 3GPP TS 23.273 V17.4.0 includes 5G Location services (LCS). With an immediate or deferred location request, LCS client expects to receive location information for a target UE or group of UES. The deferred location request is satisfied when an event occurs e.g., UE availability, UE enters an area, periodic location, UE motion. In order to determine the position of a given UE, LCS benefits from different positioning methods. The LCS clients sends the request to GMLC, then AMF and LMF (Location management Function) are triggered to determine the location of the target UE or group of UEs.

[0017] In some scenarios and deployments (e.g., industry), location estimation and prediction of devices play a crucial role in planning certain tasks.

[0018] A first problematic aspect relates to the Restricted Set of Input Data for Location Predictions: The positioning services provided by the networks rely on the data generated in the 5G system (network and UEs). As one of the closest technologies, NWDAF is capable of generating UE- related analytics and predictions (i.e., mobility, location accuracy, proximity, movement behavior) by collecting relevant information from NFs and / or GAM. However, the scope of the input data for analytics is restricted to the 3GPP specifications, which defines the type of data stored in NFs / OAM about a specific UE or a group of UEs. The input data defined by 3GPP for each of the analytic types and the data sources may not be enough to maximize the performance (e.g., accuracy) and scope (e.g., output data type) of the UE-related analytics / predictions. The 0AM and related NFs (e.g., LMF, AMF) are already capable of collecting measurements sent by the UE. For example, Enhanced Cell ID (E-CID) positioning techniques use UE measurements over a positioning protocol (e.g., LPP (LTE Positioning Protocol) or NRPP (NR positioning protocol)). A second problematic aspect relates to low performance analytics: The existing literature proposes a pre-defined set of input features for positioning related inference services. When a solution (e.g., NWDAF) provides low performance positioning prediction service, there might be a need to add new features to make the model more complex. Avoiding underfitting by adding new features may enhance the performance of the analytics / predictions with higher accuracy.

[0019] Summary

[0020] Accordingly a need exists to overcome the above-mentioned problems and to further improve the positioning prediction of a mobile entity.

[0021] This need is met by the features of the independent claims. Further aspects are described in the dependent claims.

[0022] According to a first aspect a method is provided carried out at a positioning entity configured to predict a position of the mobile entity connected to the cellular network. The positioning entity receives from the cellular network first position related data for a plurality of mobile entities including at least a first mobile entity connected to the cellular network. Furthermore, second position related data is received from the first mobile entity wherein the second position related data is not related to a communication of the first mobile entity with the cellular network. Furthermore, a future position of the first mobile entity is determined based on the first position related data and the second position related data. The future position of the first mobile entity is then reported to the cellular network for further processing at the cellular network.

[0023] Furthermore the corresponding positioning entity is provided configured to carry out the above- mentioned steps and the steps as discussed in further detail below.

[0024] Additionally, a computer program comprising program code is provided to be executed by at least one processing unit of a positioning entity, wherein execution of the program code causes the at least one processing unit to carry out a method as discussed above or as discussed in further detail below. Furthermore a carrier comprising the computer program is provided wherein the carrier is one of an electronic signal, optical signal, radio signal, and computer readable storage medium. With the use of position data not related to the communication of the first mobile entity with the cellular network itself, it is possible to improve the positioning and it is especially possible to use data collected from any source as the data used for the positioning is not restricted to the data used for the communication exchange with the network. The positioning data not related to the communication include data not generated by the cellular network for the communication or addressing of the UE in the environment of the network. It can include none 3GPP related data.

[0025] Other devices, systems, methods, features and advantages will be or will become apparent to one with skill in the art upon examination of the following detailed description and figures. It is intended that all such additional systems, methods, features and advantages be included within this description using the scope of the invention and be protected by the claim set.

[0026] Brief description of the drawings

[0027] The foregoing and additional features and effects of the application will become apparent from the following detailed description when read in conjunction with the accompanying drawings in which like reference numerals refer to like elements.

[0028] Figure 1 shows a schematic architectural overview over a system in which a positioning entity determines a future position of the mobile entity.

[0029] Figure 2 shows a schematic view of the neural network that can be used in the positioning entity for predicting the future position.

[0030] Figure 3 shows a schematic view of a message exchange between the involved entity for predicting the position of the mobile entity.

[0031] Figure 4 shows a schematic view of a possible implementation of the system of figure 1 in an asset administration shell.

[0032] Figure 5 shows a schematic view of an application of the determination of the future position in a factory. Figure 6 shows a schematic view of a flowchart including the steps carried out by the positioning entity to predict the future position of the mobile entity.

[0033] Figure 7 shows a schematic architectural view of the positioning entity configured to determine the future position of the mobile entity.

[0034] Detailed Description

[0035] In the following, embodiments of the invention will be described in detail with reference to the accompanying drawings. It is to be understood that the following description of embodiments is not to be taken in a limiting sense. The scope of the invention is not intended to be limited by the embodiments described hereinafter or by the drawings, which are to be illustrative only.

[0036] The drawings are to be regarded as being schematic representations, and elements illustrated in the drawings are not necessarily shown to scale. Rather, the various elements are represented such that their function and general purpose becomes apparent to a person skilled in the art. Any connection or coupling between functional blocks, devices, components of physical or functional units shown in the drawings and described hereinafter may also be implemented by an indirect connection or coupling. A coupling between components may be established over a wired or wireless connection. Functional blocks may be implemented in hardware, software, firmware, or a combination thereof.

[0037] Within the context of the present application, the term “mobile entity” or “user equipment” (UE) refers to a device for instance used by a person (i.e. a user) for his or her personal communication. It can be a telephone type of device, for example a telephone or a Session Initiating Protocol (SIP) or Voice over IP (VoIP) phone, cellular telephone, a mobile station, cordless phone, or a personal digital assistant type of device like laptop, notebook, notepad, tablet equipped with a wireless data connection. The UE may also be associated with nonhumans like animals, plants, or machines. In the present context, the UE is especially connected to a device providing a service for an industrial application. A UE may be equipped with a SIM (Subscriber Identity Module) or electronic-SIM comprising unique identities such as IMSI (International Mobile Subscriber Identity), TMSI (Temporary Mobile Subscriber Identity), or GUTI (Globally Unique Temporary UE Identity) associated with the user using the UE. The presence of a SIM within a UE customizes the UE uniquely with a subscription of the user. For the sake of clarity, it is noted that there is a difference but also a tight connection between a user and a subscriber. A user gets access to a network by acquiring a subscription to the network and by that becomes a subscriber within the network. The network then recognizes the subscriber (e.g. by IMSI, TMSI or GlITI or the like) and uses the associated subscription to identify related subscriber data. A user is the actual user of the UE, and the user may also be the one owning the subscription, but the user and the owner of the subscription may also be different. E.g. the subscription owner may be the parent, and the actual user of the UE could be a child of that parent. In the current application the UE is connected to a device providing a service for an industrial application.

[0038] In the following a solution is discussed which provides a high-performance positioning prediction service for cellular networks such as non-public networks, by way of example a 5G network. However it should be understood that the application is not restricted to 5G or nonpublic networks. The solution discussed below benefits from information gathered from network capable 5G capable devices such as a battery status or current task in manufacturing along with data already aggregated in the network itself. Collecting data from the end-user device related to the device state and industrial processes such as asset management, resource planning can provide valuable insights that can impact the predicted location. By way of example in industrial application planning the tasks and coordinating the robots such as automated guided vehicles, AGVs, can already know the destination coordinates for a particular device. This application layer information is normally not available in the network or the UE, but is a promising feature that enhances the positioning prediction performance such as the accuracy of the coordinates, direction, and past due to obstacles.

[0039] Figure 1 shows a schematic architectural view of a system allowing a precise prediction of a position of a mobile entity 200 connected to a cellular network 300. The mobile entity or UE 200 can be part of or can be connected to an industrial application 260 the industrial application may be a stand-alone unit or implemented within any other note may be located in a cloud infrastructure and can provide services such as planning the tasks and coordination of a robot, not shown, connected to the UE 200. Furthermore a positioning entity 100 is provided configured to determine and predict a position of the UE 200 wherein information from the industrial application, the device itself and / or from the network is used to determine the position of the UE 200 in the future. The positioning entity 100 can include a positioning analytics and prediction unit 150, a data collection unit 151 responsible for collecting data from different sources including the network and the application or device itself. A data analytics and processing unit 152 is configured to generate the statistics based on the collected and based on historical data provided to the system. A prediction unit 153 produces the prediction and may include a machine learning model which might be trained with data provided by the data collection unit 151. A reporting unit 154 can be provided for reporting the predictions to the network 300 and / or to the mobile entity 200 .

[0040] The list of data that can be collected by the positioning entity, especially the data collection unit 151 can include at least one of the following pieces of information:

[0041] -battery status,

[0042] -active or idle time of device,

[0043] -current location and destination coordinates,

[0044] -the current path and trajectory,

[0045] -the characteristics of the participated tasks in a value chain and devices in collaboration, -scheduled updates,

[0046] -radio network parameters such as network quality reception, the power used during transmission etc.

[0047] This information will be collected, processed, predicted and evaluated within the positioning entity 100.

[0048] Data collection unit 151 is responsible for the monitoring and collecting cellular based positioning information and analytics from the network, the mobile entity 200 itself and the industrial application 260. The network information can come from the AMF, GMLC or OAM. This set of information that can be collected from the network 300 can include the following pieces of information or any other information provided by the network itself. The information includes

[0049] -UE locations, associated positioning technology and accuracy,

[0050] -handover operations, MWDAF (Network Data analytics Function) analytics and predictions related to the UE such as the mobility, movement behavior and proximity,

[0051] -UE moving direction and velocity, -timestamp.

[0052] Furthermore, the data collection unit 151 collects data from the device and the industrial application related to its context. The data collection process may be executed periodically to grab the changes and in order to generate the time series data. In order to collect the relevant information from the network, device and industrial application, the data collection unit 151 needs to utilize the existing exposure capabilities provided by the system. The standardized exposures provided by the network such as a SEAL or NEF or any proprietary interface can be used to retrieve the positioning related information and analytics provided by the cellular network 300. A similar approach can be used to fetch the data from the mobile entity 200 and industrial application 260. If supported the positioning entity or here the data collection unit 151 can subscribe to periodic or event based exposures provided. The list of data that can be collected by the devices and / or the industrial application can include the following nonlimiting list:

[0053] -battery status

[0054] -current location and destination coordinates

[0055] -the current path and the trajectory

[0056] -participated tasks in a value chain and devices in collaboration

[0057] -scheduled updates

[0058] -radio network parameters including the network quality reception, the power used during transmission etc.

[0059] T o fetch relevant data from the mobile entity 200 or the industrial application the data collection unit 151 can benefit from any proprietary interface and application programming interface, API, set provided by the device or the industrial application itself.

[0060] The data analytics and processing unit 152 can provide the following:

[0061] This entity can be in charge of processing the collected data, filtering and ensuring the capability between the retrieved information. Besides, data integration is done by combining the data collected from different sources such as the network and the device itself based on common identifiers like device IDs and timestamps.

[0062] This unit 152 is integrated with relevant methods to provide descriptive statistics by summarizing the historical data and generating insights. By using such techniques, this unit 152 can generate knowledge such as mean number of UEs in a cell and the most crowded location during the day. In addition to this capability, this unit 152 can accommodate techniques providing inferential statistics that uses the data and draw conclusions. Based on a given sample, inferential statistic methods such as analysis of variance (ANOVA) in St, L., & Wold, S. (1989): “Analysis of variance (ANOVA). Chemometrics and intelligent laboratory systems”, 6(4), 259-272, can be used to make generalizations about a larger population. For example, the movement of UEs in different groups (e.g., UEs consuming different services) can be further analyzed to see if there is a statistically significant differences among groups. This collection of information can be useful for further planning in the business and network.

[0063] Detection of the correlations and patterns within collected data that could be beneficial for location prediction is done in prediction unit 153. Afterward, an appropriate machine learning model(s) for location prediction can be chosen. Based on the selected ML algorithm, training, validation and evaluation (Mean Absolute Error or Root Mean Square Error) of the model to assess the accuracy of the model will be executed. Based on the requirement, a linear model (e.g., linear regression) can be designed to predict the location coordinates for a given UE. On the other hand, the complex relations between the input features can be determined through a non-linear model (e.g., neural network). An example of neural network 500 with multiple hidden layers 520 such as hidden layers 521-523 is depicted in Error! Reference source not found., where the input features is denoted as X, and represents the mthfeature of ithexample.

[0064] As observed, the input feature set includes the ones collected from network, devices and / or industrial applications, where the output node or layer 530 represents the predicted location coordinates. The number of hidden nodes and layers can be optimized to maximize the accuracy or minimize the error.

[0065] Depending on the input set, it is preferred to continuously evaluate the model, assess the accuracy and implement a mechanism for continuous improvement. Since the proposed solution is capable of collecting various data types from different sources, it is able to determine the most relevant features to predict the location and maximize the accuracy accordingly. In this direction, the prediction module may be integrated with feature selection techniques, so that a subset of available features in the original set can be picked, reduce the feature space dimension and accelerate the training process by improving the accuracy. This process can be executed offline, while the original model continues running inference on the run. If accuracy is improved with the arrival of new data set, the original model can be replaced with the one trained and tested offline. As an example, if it is concluded that battery status of the end-user device is not predictive of the estimated location, it can be removed from the set of input features used for the prediction model to reduce the complexity and enhance the performance. Since the existing technology proposes to use models working with a pre-defined set of input features, which are collected from functions in the network, the proposed solution here extends the scope so that different type of data can be collected from multiple sources. The useful and meaningful information collected from the mobile entity 200 or the application 260 itself enhances the prediction accuracy since some data types are already correlated with the positioning, but the network 300 is not able to fetch them. Therefore, the solution proposed here enables interoperability across domains (e.g., industrial applications, network) and achieves data exchange between different systems. For example, it is assumed that an industrial robot (e.g., AGV) participates in some manufacturing related tasks by collaborating with the other devices. The industrial applications already know the task details, participating devices, the location of the production line, the obstacles, surroundings and many other information. This set of information is directly related to the location of the devices at the factory floor. However, the existing technology in the network used for mobility related predictions as known in the art before were not able to collect this knowledge. The proposed prediction model gathers various data types from multiple sources that increase the prediction performance. Additionally, by continuously monitoring its own performance, the prediction model can be modified accordingly or retrained to achieve higher levels of performance.

[0066] The unit 154 reporting feedbacks can be based on the measurements and computations, the confidence and accuracy of prediction as well as position of the device at t+1 , will be reported as feedback to network 300, industrial app 260, mobile entity 200 and / or other devices. The feedback provided to the network can be used for multiple purposes such as resource allocation, fault management, proactive configuration management and energy efficiency. There might be other benefits of providing feedback to the network.

[0067] In addition to the network, end-user mobile entity 200 or industrial application 260 can benefit from the generated feedback. Based on the predicted location, the path of the device can be optimized (e.g., obstacles on the path), asset management processes can be optimized, energy saving can be activated, and manufacturing tasks can be planned accordingly.

[0068] Error! Reference source not found. Error! Reference source not found, depicts a sequence diagram of the proposed solution. The positioning entity 100 in step S20 initiates a data collection phase by requesting the 5G network about the position and analytic reports at time tj ( with the corresponding response from the network in steps S21). In step S S22, the device mobility is requested in step S22 with the response from the network in step S23. In step S22 the positioning entity requests the position at t+1 predicted by the network. The positioning entity uses the predicted location generated by the network to make its own prediction at t+1. Then it gathers the contextual data from the mobile entity 200 or the application in steps S25 and S26. For simplification only the Mobile entity 200 is shown in Error! Reference source not found. Error! Reference source not found.. Nevertheless, similarly to the mobile entity 200 the industrial application 260 could also provide this information. Moreover, they could both be combined and interface to the Positioning entity 100 at the same time. Thereafter, having all the necessary information by optionally utilizing ML algorithms the location of the devices at t(i+1) will be predicted by analyzing, validating, and calculating the position accuracy (step S27, S28). Finally, the reports in the shape of feedbacks and updates will be disseminated to network in S29 and optionally also to the mobile entity 200 in a step not shown.

[0069] As shown in Error! Reference source not found, the positioning entity 100 and all accommodated functions can be implemented with Asset Administration Shell (AAS). AAS is an industrial specification that creates the digital representation of assets. The standardized communication and interfaces defined by AAS enables the interoperability between different systems with different communication models.

[0070] It is proposed to create the AAS of 5G network, named 5G Network AAS 400, including the positioning entity 100. AAS, in principle, includes two parts: (i) passive, (ii) active. The passive part 420 contains the data and information provided by the asset and other AAS instances. On the other hand, active part 410 implements the digital twin-like capabilities in AAS where any decision-making algorithm or analytical models can be implemented.

[0071] The positioning related data exposed by the network 300 can be stored in 5G Network AAS. In order to have an interoperable communication with the industrial devices by eliminating vendor dependency and collect contextual data for prediction and analytics purposes, AAS of the devices can be implemented (i.e., 5G UE AAS 201). So, by using device proprietary interfaces and APIs, the contextual data of each end-user device can be stored in its own AAS, which can then be communicated with the 5G Network AAS 400 through standard language and communication model. Data collected from different sources can be stored in AAS, in form of submodels. The data collection, data analytics & processing, prediction and feedback entities (positioning entity 100) can be implemented in the active part 410 of 5G Network AAS 400. When needed, the active part functions can access the passive part 420, write, read or modify the content in the specific sub-model accordingly. The examples of the sub-models can be localization, positioning analytics and alarm / events that can be used to store various kind of information.

[0072] In connection with Fig. 5 a use case, a more concrete example scenario is explained in order to clarify the idea. The use case includes a smart factory environment equipped with a 5G NPN (non-public network) and 5G enabled devices. A smart 5G enabled AGV (Automated Guided Vehicle) 600 is considered as the specific UE to be examined during this use case. It is assumed that UE is either embedded in the AGV 600 or connected via a separate connection. The aim is to enhance the position prediction performance (e.g., accuracy) of the UE using contextual information, that is gathered from the device with the help of AAS.

[0073] The AGV 600 is moving at the trajectory shown in Fig. 5Error! Reference source not found.. Without having detailed contextual information, current prediction methods can only make predictions based on the historical position data, which will stochastically predict whether the device will move towards routes A, B or C. This prediction may not be very precise because in every mission, which direction the AGV will move is unknown.

[0074] After applying the proposed method, according to AGV’s participated tasks in value chain and devices in collaboration, planned operation route of the AGV 600 is provided as the contextual information to be used to support enhanced position prediction. This information will be transferred to the proposed Positioning entity 100, e.g. via AAS. Other information modeling frameworks such as Open Platform Communications- Unified Architecture (OPC-UA), digital twin representations, APIs and other standardized or proprietary solutions providing exposure capabilities from the 5G device or the industrial application (providing the planned route) may also provide context information. The AGV 600 has already secure connection with its own AAS (i.e., 5G UE AAS), via the proprietary interface. The information about the planned operation route, along with other relevant set of data, can be exposed by AGV 600 to its AAS periodically. Then, AAS of the AGV can exchange this information with the AAS of 5G Network (i.e., 5G Network AAS). This information and knowledge exchange can be realized through the standardized AAS communication model and interfaces. It should be noted that 5G AAS is capable of storing the device contextual information collected from different devices, through their AAS representations.

[0075] With this information the path that the AGV 600 will follow is now known by the Positioning entity 100 embedded in the 5G AAS. Together with other position related information, such as AGV velocity, acceleration, heading, current position, current time, elapsed time etc., the prediction module will calculate the future position of the AGV 600 to be on top of the given path. To give a concrete example, suppose that the planned path of AGV is through Route C as shown in Error! Reference source not found.. Given that this information is provided to the Positioning entity 100 as an input via AAS interaction, now the function will not take other possible routes, Route A and Route B into consideration. Since the planned route gathered from the device itself has a strong impact on the predicted location and is one of the important descriptors, using this information as an input feature of the prediction model will increase the probability of generating the predicted location that is on Route C. It will calculate the position prediction such that the predicted position will be one of the locations on top of the Route C, which is the correct path. This will drastically increase the possibility of having more accurate position predictions unless the planned operation is not followed due to an extraordinary event such as failures etc.

[0076] Moreover, an industrial warehouse scenario can be assumed in which human and robot would like to work in a safe environment. In these scenarios, the position of the objects in the warehouse is not fixed and can be changed depending on the warehouse stock. Generally, to ensure the safety of humans, robots are contained in a separate fenced section, which makes the human robot collaboration (HRC) almost impractical.

[0077] Therefore, to ensure a safe environment and utilize the benefits of HRC at the same time is very challenging unless the position and path of the robots and humans can be made sure and exchanged using the embodiments discussed above.

[0078] In connection with figure 6 some of the steps carried out by the positioning entity 100 are summarized. In step S 111 first position related data are received for a plurality of mobile entities wherein the first data included the data from the first mobile entity 200 connected to the cellular network. This was also discussed above in more detail in connection with step S 21. Furthermore, from the mobile entity the second position related data are received wherein the second position related data are not related to a communication of the first mobile entity with the network. This step S112 was discussed above in connection with step S 26. In step S113 a future position of the first mobile entity 200 is determined based on the first position related data and the second position related data as also discussed above in connection with step S 27. The future position of the first mobile entity is then reported in step S 114 to the cellular network for further processing at the cellular network as also discussed in connection with S29 above.

[0079] In connection with figure 7 a further schematic architectural view of the positioning entity is shown which can determine the future position of the mobile entity as discussed above. The positioning entity 100 comprises an interface 110 used for transmitting user data or control messages to other entities and used for receiving user data or control messages from other entities. The interface 110 is especially qualified to receive the different position related data from the network or the mobile entity. The entity 100 furthermore comprises a processing unit 120 which is responsible for the operation of the positioning entity 100. The processing unit 120 can comprise one or more processors and can carry out instructions stored on a memory 130, wherein the memory may include a read-only memory, a random access memory, a mass storage, a hard disk or the like. The memory 130 can furthermore include suitable program code to be executed by the processing unit 120 so as to implement the above-described functionalities where the positioning entity 100 is involved.

[0080] From the above said some general conclusions can be drawn.

[0081] According to one aspect the first mobile entity can be connected to a first device such as the robot providing an industrial application 260 and the second position related data received from the mobile entity can include the application related data related to the industrial application.

[0082] The second position unit data received from the mobile entity can comprise information such as Battery status of the first mobile entity, a current location of the first mobile entity, destination coordinates of the first mobile entity, a planned trajectory of a path to be used by the first mobile entity, a task to be carried out as a service for the industrial application, information about at least one second device with which the first mobile entity cooperate in providing the service for the industrial application. Together with the first position related data from the network the prediction of the position of the mobile entity can be improved. It is possible to use a trained machine learning module such as the neural network 500 shown in figure 2 is used which receives as input at least the first position related data entity second position related data and provides as an output the future position of the first mobile entity.

[0083] The positioning entity may further carry out a statistical analysis of the first and second position related data in order to generate statistically filtered data entity and the position may then be determined taking the count the statistically filtered data.

[0084] The statistically filtered data may be used as further input to the trained machine learning module as also shown in figure 2.

[0085] Furthermore it is possible that the future position of the first mobile entity is transmitted and reported to the first mobile entity 200 for a further use at the first mobile entity.

[0086] The cellular network 300 may be a non-public network, however also a public network may be used instead of the non-public network.

[0087] The first mobile entity may be connected to an automated guided vehicle 600 or any robot and the future position is the position of the automatic guided vehicle or the robot. The second position related data can include a planned route of the automatic guided vehicle and the future position is determined taking into account the planned route of the guided vehicle as it was provided by the second position related data.

[0088] Furthermore it is possible that the determining and reporting of the position is implemented in an asset administration shell environment.

[0089] The solution discussed above enables integration of different systems such as industrial processes and networks and it can implement the capability of collecting data from any source. It removes the constraint of using predefined set of input features for positioning analytics and predictions and thus adds flexibility by adding more features. Furthermore the most relevant features used for the prediction can be identified to improve the performance.

[0090] The positioning prediction service discussed above can be used for proactive resource management in the network. To enhance and optimize the resource utilization and provide service assurance for the mobile entity the estimated location may be used accordingly. The proposed solution can also be useful for estimating the location when there is no connectivity. The estimated location can be reported to the device when out of coverage so that it can be used for further operation such as asset management, optimizing the path planning etc.

[0091] The estimated locations can also be beneficial for enhancing the energy efficiency for both the mobile entity and the network. By periodically monitoring the alternative routes and predicted location of all devices, the most energy efficient path can be selected for each mobile entity. Similarly, based on the predicted locations the energy efficiency can be increased in the network as well. Based on the predicted locations an expected load in certain areas a particular set of cells or a transmit and / or receive antenna arrays can be switched off to increase energy savings.

Claims

Claims1 . A method carried out at a positioning entity (100) configured to predict a position of a mobile entity connected to a cellular network (300), the method comprising:- receiving (S21 S111), from the cellular network, first position related data for a plurality of mobile entities including at least a first mobile entity (200) connected to the cellular network, -receiving (S26, S112), from the first mobile entity (200), second position related data, the second position related data not being related to a communication of the first mobile entity (200) with the cellular network (300),- determining (S27, S113) a future position of the first mobile entity (200) based on the first position related data and the second position related data,- reporting (S29, S114) the future position of the first mobile entity to the cellular network for further processing at the cellular network.

2. The method of claim 1 , wherein the first mobile entity is connected to a first device (210) providing a service for an industrial application, wherein the second position related data include application related data related to the industrial application.

3. The method of claim 1 or 2, wherein the second position related data comprise at least one of the following:- a battery status of the first mobile entity,- a current location of the first mobile entity,- destination coordinates of the first mobile entity,- a planned trajectory of a path to be used by the first mobile entity,- a task to be carried out as a service for the industrial application,- information about at least one second device with which the first mobile entity (200) and first device cooperate when providing the service for the industrial application.

4. The method of any preceding claim, wherein for determining the future position a trained machine learning module is used receiving as input at least the first position related data and the second position related data and providing as output the future position of the first mobile entity.

5. The method of any preceding claim, further carrying out a statistical analysis of the first and second position related data in order to generate statistically filtered data, wherein the future position is determined taking into account the statistically filtered data.

6. The method of claim 4 and 5, wherein the statistically filtered data is used as further input to the trained machine learning module.

7. The method of any preceding claim, further reporting the future position of the first mobile entity (200) to the first mobile entity for a further processing at the first mobile entity.

8. The method of any preceding claim, wherein the cellular network is a non-public network.

9. The method of any preceding claim, wherein the first mobile entity (200) is connected to an automated guided vehicle, wherein the future position is the future position of the automated guided vehicle.

10. The method of claim 9, wherein the second position related data include a planned route of the automated guided vehicle, wherein the future position is determined taking into account the planned route of the automated guided vehicle.

11. The method of any preceding claim wherein the determining and reporting of the future position is implemented in an Asset Administration Shell environment.

12. A positioning entity configured to predict a position of a mobile entity connected to a cellular network, the positioning entity being configured to:- receive (S21 S111), from the cellular network, first position related data for a plurality of mobile entities including at least a first mobile entity (200) connected to the cellular network, -receive (S26, S112), from the first mobile entity (200), second position related data, the second position related data not being related to a communication of the first mobile entity (200) with the cellular network (300),- determine (S27, S113) a future position of the first mobile entity (200) based on the first position related data and the second position related data,- report (S29, S114) the future position of the first mobile entity to the cellular network for further processing at the cellular network.

13. The positioning entity of claim 12, wherein the first mobile entity is connected to a first device (210) providing a service for an industrial application, wherein the second position related data include application related data related to the industrial application.

14. The positioning entity of claim 12 or 13, wherein the second position related data comprise at least one of the following:- a battery status of the first mobile entity,- a current location of the first mobile entity,- destination coordinates of the first mobile entity,- a planned trajectory of a path to be used by the first mobile entity,- a task to be carried out as a service for the industrial application,- information about at least one second device with which the first mobile entity (200) and first device cooperate when providing the service for the industrial application.

15. The positioning entity of any of claims 12 to 14, further being configured, for determining the future position, to use a trained machine learning module receiving as input at least the first position related data and the second position related data and providing as output the future position of the first mobile entity.

16. The positioning entity of any of claims 12 to 15, further being configured to carry out a statistical analysis of the first and second position related data in order to generate statistically filtered data, and to determine the future position taking into account the statistically filtered data.

17. The positioning entity of claim 15 or 16, further being configured to use the statistically filtered data as further input to the trained machine learning module.

18. The positioning entity of any of claims 12 to 17, further being configured to report the future position of the first mobile entity (200) to the first mobile entity for a further processing at the first mobile entity.

19. The positioning entity of any of claims 12 to 18, wherein the cellular network is a non-public network.

20. The positioning entity of any of claims 12 to 19, wherein the first mobile entity (200) is connected to an automated guided vehicle, wherein the future position is the future position of the automated guided vehicle.

21. The positioning entity of claim 20, wherein the second position related data include a planned route of the automated guided vehicle, the positioning entitO being configured to determine the future position taking into account the planned route of the automated guided vehicle.

22. The positioning entity of any of claims 12 to 21 , further being configured to implement the determining and reporting of the future position in an Asset Administration Shell environment.

23. A computer program comprising program code to be executed by at least one processing unit of a positioning entity, wherein execution of the program code causes the at least one processing unit to carry out a method as mentioned in any of claims 1 to11.

24. A carrier comprising the computer program of claim 23, wherein the carrier is one of an electronic signal, optical signal, radio signal, and computer readable storage medium.

Images