Abstract: A search system provides search results with images of products based on associations of primary products and secondary products from product image sets. The search system analyzes a product image set containing multiple images to determine a primary product and secondary products. Information associating the primary and secondary products are stored in a search index. When the search system receives a query image containing a search product, the search index is queried using the search product to identify search result images based on associations of products in the search index, and the result images are provided as a response to the query image.

Abstract: A method performed by a node for predicting a behavior of users of a communications network is described. The node manages an artificial neural network. The node merges a first pre-existing predictive model of the behavior in a first group of users with a second model of the behavior in a second group of users. The merging comprises establishing connections between the first model and the second model. Each of the connections has a respective weight. The respective weights of the connections are learned by respective connections of neurons in the artificial neural network based on data from a third group of users. The node also obtains a third model for predicting the behavior in the third group of users, based on the merged models and the data from the third group of users.

Abstract: It is provided a method for adjusting alignment for microwave transmissions from a microwave transmitter to a microwave receiver based on a reinforcement learning, RL, model. The method comprises the steps of: obtaining state space comprising external state space and internal state space, the external state space comprising at least one value of a parameter related to environmental conditions, and the internal state space relates to alignment of the microwave transmitter; determining an action in an action space, the action space comprising actions to adjust alignment of the microwave transmitter; obtaining a measurement of path loss for a transmission from the microwave transmitter to the microwave receiver; determining a reward value based on the path loss, wherein an increase in path loss results in a reduced reward value; and adjusting the RL model based on the obtained state space, the determined action and the determined reward value.

Abstract: Managing network event data in a telecommunication network. The method includes obtaining a plurality of datasets having network event data associated with products provided by a different vendor, the network event data comprising a plurality of data instances representing a plurality of network event features. The method further comprises obtaining metadata describing the network event data in the datasets and generating feature description vectors from the obtained metadata, calculating a metric of the relation between network event features represented in the datasets and constructing a graph of the network event features represented in the datasets, with edges weighted according to the calculated metric, partitioning the graph into clusters by minimising an edge cut between network event features, labelling network event features in the cluster with a normalised network event feature index and inputting the labelled network event data to a model for making at least one of recommendations or predictions.

Abstract: (57) Abstract: A method, performed by a first node (111), for handling parameters to configure a second node (112). The first node (111) determines (207) parameters to configure the second node (112). The determining (207) is based on an analysis by a Generative Adversarial Network comprising performing iteratively: i) generating a set of parameters estimated to be linked to a performance of a first group of nodes (121) being above a threshold, and ii) discriminating between the generated set of parameters and a first set of parameters observed to be linked to the performance of the first group of nodes (121) being above the threshold to obtain a score for every parameter. The score indicates how different the generated parameters and the first set of parameters are. The determined parameters have the score resulting from the discriminating being lower than anotherthreshold. The first node (111) also outputs an indication comprising the determined parameters.

Abstract: Methods and devices optimize backup power management for network devices connected to backup power sources including two different batteries. Machine learning techniques are used to predict upcoming power outages affecting the network device based on power-related historic information and current conditions. A backup-power operation plan prescribing usage of the backup power sources during the predicted power outages is then generated. The backup-power operation plan schedules a battery among the backup power devices to be used at least twice without being recharged.

Abstract: A compound or complex containing a rare earth element is impinged with a pulsed laser that is so controlled as to photochemically reduce and obtain a rare earth metal (REM). A mixture of REM salts can be impinged using laser light tuned to selectively reduce a particular rare earth-containing salt of the mixture to separate out as its respective rare earth metal.

Abstract: An efficient and green method for selective extraction of Li from end-of-life secondary LIBs of any capacity and size is provided. Electrochemical driven selective lithium deposition is targeted at the anode/separator interface of the end-of-life LIB. The deposited Li is recovered by processing of an opened or dismantled battery using only distilled or de-ionized water. The process not only enables the recovery of the plated lithium at the anode/separator interface, but also extracts the lithium from the organic salts and/or inorganic salts in the solid electrolyte interface (SEI) layers and from the electrolyte in the separator. In addition, the method partially strips the cyclable Li from the cathode and concentrates it at the anode/separator interface. The concentrated Li is extracted by using aqueous solution such as distilled or de-ionized water followed by recovery of the Li from aqueous solution. After Li recovery from the anode, the method can also enable the recovery of battery-grade graphite.

Abstract: Methods and devices optimize backup power management for network devices connected to backup power sources including two different batteries. Machine learning techniques are used to predict upcoming power outages affecting the network device based on power-related historic information and current conditions. A backup-power operation plan prescribing usage of the backup power sources during the predicted power outages is then generated. The backup-power operation plan schedules a battery among the backup power devices to be used at least twice without being recharged.

Abstract: Fast charging of a lithium-ion battery at 4C rate or 5C rate or more is improved by applying an external magnetic field relative to the battery to establish magnetic field lines that extend in a direction of primary movement of lithium ions toward the graphite anode during fast charging. Deleterious degradation of the graphite anode from repeated fast charging can be reduced or eliminated by practice of the invention.

Abstract: Managing network event data in a telecommunication network. The method includes obtaining a plurality of datasets having network event data associated with products provided by a different vendor, the network event data comprising a plurality of data instances representing a plurality of network event features. The method further comprises obtaining metadata describing the network event data in the datasets and generating feature description vectors from the obtained metadata, calculating a metric of the relation between network event features represented in the datasets and constructing a graph of the network event features represented in the datasets, with edges weighted according to the calculated metric, partitioning the graph into clusters by minimising an edge cut between network event features , labelling network event features in the cluster with a normalised network event feature index and inputting the labelled network event data to a model for making at least one of recommendations or predictions.

Abstract: Methods and devices optimize backup power management for network devices connected to backup power sources including two different batteries. Machine learning techniques are used to predict upcoming power outages affecting the network device based on power-related historic information and current conditions. A backup-power operation plan prescribing usage of the backup power sources during the predicted power outages is then generated. The backup-power operation plan schedules a battery among the backup power devices to be used at least twice without being recharged.

Abstract: It is provided a method for adjusting alignment for microwave transmissions from a microwave transmitter to a microwave receiver based on a reinforcement learning, RL, model. The method comprises the steps of: obtaining state space comprising external state space and internal state space, the external state space comprising at least one value of a parameter related to environmental conditions, and the internal state space relates to alignment of the microwave transmitter; determining an action in an action space, the action space comprising actions to adjust alignment of the microwave transmitter; obtaining a measurement of path loss for a transmission from the microwave transmitter to the microwave receiver; determining a reward value based on the path loss, wherein an increase in path loss results in a reduced reward value; and adjusting the RL model based on the obtained state space, the determined action and the determined reward value.

Abstract: The disclosure provides methods, apparatus and machine-readable mediums for mitigating power failures in base stations in wireless communication networks. In one aspect of the disclosure, a method for mitigating power failures in a wireless communication network is provided. The method comprises, responsive to a determination that a primary power supply for a base station in the wireless communication network has failed or is predicted to fail, initiating deployment of an unmanned aerial vehicle to the base station to charge one or more batteries providing a secondary power supply for the base station.

Abstract: A method (100) for managing a group of Unmanned Aerial Vehicles (UAVs) operable to fly in a formation is disclosed, each UAV being programmed with a task to be performed by the UAV. The method is performed in a controller UAV of the group and comprises receiving UAV status information from UAVs in the group (110), obtaining information on a current formation of the group (120) and combining the received UAV status information with the information on current group formation to form a representation of a first state of the group (130). The method further comprises using a machine learning model to predict, on the basis of the first state of the group, an optimal formation transition to a new formation (140) and to instruct the UAVs in the group to perform the predicted optimal formation transition (150). An optimal formation transition is a transition to a formation that will minimise predicted total energy consumption for all UAVs in the group to complete their tasks.

Abstract: Embodiments of the present invention provide an electromagnet alignment system for AM or 3D printing technology providing improved in-situ alignment of the magnetic particulate material as it is dispensed during deposition to form a 3D shape. In-situ alignment of the magnetic particulate material can be controlled to be unidirectional or multi-directional.

Abstract: Methods and devices optimize backup power management for network devices connected to backup power sources including two different batteries. Machine learning techniques are used to predict upcoming power outages affecting the network device based on power-related historic information and current conditions. A backup-power operation plan prescribing usage of the backup power sources during the predicted power outages is then generated. The backup-power operation plan schedules a battery among the backup power devices to be used at least twice without being recharged.

Abstract: The present disclosure relates to a method performed by a processing device (4) of a Work Force Management (WFM) system (1) for predicting timely completion of a work order for service at a remote site (3) of a telecommunication system (2). The method comprises obtaining information about a time period allowed for completion of the work order. The method also comprises selecting, based on previous knowledge about the remote site, a set of factors which may affect service at the remote site. The method also comprises updating information about the selected factors. The method also comprises, at a first point in time within the allowed time period, obtaining information about a current status of the work order. The method also comprises, based on the updated information and the work order status at said first point of time, predicting whether it is likely that the work order will be completed within the allowed time period.