Abstract: Novel techniques and mechanisms enable processing of heavy deep learning workloads on standard edge network devices to optimize the overall inference throughput of the network while meeting Service Level Agreement(s) (SLAs). The techniques can include receiving a deep learning model, determining a graph structure of the deep learning model including neurons organized in layers (the layers including an input layer, a plurality of hidden layers, and an output layer), assigning to a first IP subnet, at least a part of a first hidden layer of the plurality of hidden layers, assigning to a second IP subnet, at least a part of a second hidden layer of the plurality of hidden layers, and deploying the parts of the first and second hidden layers to edge devices as containerized applications with assigned IP addresses, which may be hidden from the user and/or third party application.

Abstract: A device for testing the contents of a switch cabinet following installation of equipment components (2a-2e) according to a plan and the wiring thereof via terminals (6) provided therefor by means of designated electrical cables (5), includes a camera unit (7) for the at least two-dimensional image capture of the components (2a-2e) fully installed and wired on an installation board (4) of a switch cabinet (1). The device can further include an evaluation unit (8) for carrying out a comparison between an imaged layout (100), on the basis of the image capture, of the actual state with a provided plan layout (200) of the target state.

Abstract: Novel techniques and mechanisms enable processing of heavy deep learning workloads on standard edge network devices to optimize the overall inference throughput of the network while meeting Service Level Agreement(s) (SLAs). The techniques can include receiving a deep learning model, determining a graph structure of the deep learning model including neurons organized in layers (the layers including an input layer, a plurality of hidden layers, and an output layer), assigning to a first IP subnet, at least a part of a first hidden layer of the plurality of hidden layers, assigning to a second IP subnet, at least a part of a second hidden layer of the plurality of hidden layers, and deploying the parts of the first and second hidden layers to edge devices as containerized applications with assigned IP addresses, which may be hidden from the user and/or third party application.

Abstract: A DC-to-DC converter for an electrical aircraft propulsion system designed to be connected in series with an electrical energy storage unit of the electrical propulsion system. The DC-to-DC converter includes an inverter, a transformer and a rectifier, and further includes a current source that is connected to the rectifier and is configured to control the power passing through the DC-to-DC converter. The transformer includes a primary and two secondaries, the two secondaries sharing a common terminal designed to be connected to a high-voltage DC bus of the electrical propulsion system and two other terminals that are connected to the rectifier. The rectifier includes two arms including at least two transistors that are each in series and are connected, on the one hand, to the two other terminals of the transformer and, on the other hand, to the current source.

Abstract: This disclosure describes techniques and mechanisms for enabling a user and third party applications to dynamically partition and place heavy deep learning workloads on standard edge networks to optimize the overall inference throughput of the network while meeting Service Level Objective(s) (SLOs). The techniques may include profiling, partitioning, and splitting of the deep learning workloads, which may be hidden from the user and/or third party application. The user may user interact with a pre-deployed service through a simple SDK that resembles those used for hardware acceleration, such that the current techniques may be easily inserted into their code.

Abstract: A system and method for mounting a modular switch cabinet equipment in a switch cabinet housing (6) includes a computer-aided assistance unit (1) for determining a production-efficient assembly step sequence. The system and method further include at least one display unit (11-11?) for image and/or text information, which is installed at the assembly site for at least one fitter (M) in order to visualize manual assembly steps (A; B; D) of the determined assembly step sequence. The system and method further includes at least one input unit (12-12?) for acknowledgement of the completed assembly step (A; B; D) by the fitter (M).

Abstract: A method and a system for collecting and processing of life cycle information about a switch cabinet (1) of an industrial plant, with a documentation server (2) for linking data sheet information (A) of the built-in modules (9a-9c) installed in the switch cabinet (1) on the basis of the planned construction (4) with function information (B) about their use in the context of the switch cabinet equipment, in order to store this information in a central archive database (3) under an individual switch cabinet identification (8) in data records, wherein test and acceptance information (C) of the switch cabinet (1) is added to the data records of the central archive database (3) via the documentation server (2).

Abstract: This disclosure describes techniques and mechanisms for enabling a user to run heavy deep learning workloads on standard edge networks without off-loading computation to a cloud, leveraging the available edge computing resources, and efficiently partitioning and distributing a Deep Neural Network (DNN) over a network. The techniques enable the user to split a workload into multiple parts and process the workload on a set of smaller, less capable compute nodes in a distributed manner, without compromising on performance, and while meeting a Service Level Objective (SLO).

Abstract: A hybrid/electric propulsion architecture for a multi-rotor rotary wing aircraft, including an electricity generator driven by an internal combustion engine, and configured to operate in motor mode, a rectifier configured to convert an alternating current delivered by the electricity generator into direct current, an electrical network including a high voltage direct current (HVDC) bus, electrical energy storage means connected to the electrical network, during electrical energy regeneration on the HVDC bus, depending on the state of charge of the storage means: the storage means are configured to recover electrical energy, the storage means and the rectifier are configured to recover electrical energy, and the electricity generator operating in motor mode is configured to recover electrical energy.

Abstract: Techniques and mechanisms for automatically identifying counters/features of a network component that are related to a state change (or event) for the network component or for the network itself. For example, using data obtained from the network component around a time of the state change, delta-averages for the counters/features around the time of the state change may be determined. The delta-averages may be utilized to determine which counters/features are most descriptive for a particular state change. Determining which counters/features are most descriptive may also include determining which counters/features are most relevant, i.e., counters/features that contribute most to preserving the manifold structure of the original data or counters/features with the highest or lowest correlation with the other counters/features in the data set.

Abstract: Aviation turbine blade (10) extending in the radial direction from a blade root (14) as far as an upper partition wall (26), said blade (10) comprising a plurality of inner cavities (C1-C10) defining at least one cooling circuit, each of said inner cavities (C1-C10) being defined by walls among inner walls (40, 42), a lower surface wall (22), an upper surface wall (24), the blade root (14) and the upper partition wall (26), said blade (10) being characterized in that it comprises at least one reinforcing beam (50, 60) disposed inside one of the inner cavities (C3, C8), and connecting the blade root (14) to the upper partition wall (26), said reinforcing beam (50, 60) is not connected to the inner walls (40, 42), the lower surface wall (22) and the upper surface wall (24).

Abstract: Blade for a turbine, comprising a blade root and an airfoil (13) extending radially outwards from the blade root (12), the airfoil (13) comprising a first internal cooling circuit including an intrados cavity (33, 36) extending radially along the intrados wall (16) and a first inner wall (47, 45) arranged between the intrados wall (16) and the extrados wall (18), an extrados cavity (34, 37) extending radially along the extrados wall (18) and a second inner wall (47, 43) arranged between the intrados wall (16) and the extrados wall (18). The first cooling circuit includes one inner through cavity (35, 38) defined between two through walls (59, 57, 55, 53) each extending between the intrados wall (16) and the extrados wall (18). The intrados cavity (33, 36), the extrados cavity (34, 37) and the inner through cavity (35, 38) are fluidly connected in series.

Abstract: Techniques and mechanisms for automatically identifying counters/features of a network component that are related to a state change (or event) for the network component or for the network itself. For example, using data obtained from the network component around a time of the state change, delta averages for the features around the time of the state change may be determined. The delta averages may be utilized to determine which counters/features are most descriptive for a particular state change. The counter/features that are the most descriptive for a particular state change is as important as the change detection itself. This is especially true since in a case of an event/state change occurring, a large amount of counters/features may react to the state change or event. Thus, the techniques described herein provide for an approach to distill which counters/features contribute the most to a particular state change from a data driven perspective.

Abstract: An aviation turbine blade (10) includes at least a first lower surface cavity (C2) and a first upper surface cavity (C3), each adjacent to a first through-cavity (C1) and a second through-cavity (C4), the first upper surface cavity (C3) being adjacent to the upper surface wall (24), the first lower surface cavity (C2) being adjacent to the lower surface wall (22), each of the first and second through-cavities (C1, C4) extending from the lower surface wall (22) as far as the upper surface wall (24), the second through-cavity (C4) including a first inner wall (P1) extending from the upper surface wall (24) as far as the first through-cavity (C1), and a second inner wall (P2) extending from the lower surface wall (22) as far as the first through-cavity (C1). The first (P1) and second (P2) inner walls are not connected.

Abstract: Techniques and mechanisms for automatically identifying counters/features of a network component that are related to a state change (or event) for the network component or for the network itself. For example, using data obtained from the network component around a time of the state change, delta averages for the features around the time of the state change may be determined. The delta averages may be utilized to determine which counters/features are most descriptive for a particular state change. The counter/features that are the most descriptive for a particular state change is as important as the change detection itself. This is especially true since in a case of an event/state change occurring, a large amount of counters/features may react to the state change or event. Thus, the techniques described herein provide for an approach to distill which counters/features contribute the most to a particular state change from a data driven perspective.

Abstract: Techniques and mechanisms for automatically identifying counters/features of a network component that are related to a state change (or event) for the network component or for the network itself. For example, using data obtained from the network component around a time of the state change, delta-averages for the counters/features around the time of the state change may be determined. The delta-averages may be utilized to determine which counters/features are most descriptive for a particular state change. Determining which counters/features are most descriptive may also include determining which counters/features are most relevant, i.e., counters/features that contribute most to preserving the manifold structure of the original data or counters/features with the highest or lowest correlation with the other counters/features in the data set.

Abstract: An aviation turbine blade extending in the radial direction and presenting a pressure side and a suction side, including a plurality of pressure side cavities extending radially at the pressure side of the blade, a plurality of suction side cavities extending radially at the suction side of the blade, and at least one central cavity located in the central portion of the blade and surrounded by pressure side cavities and by suction side cavities, the blade further including a plurality of cooling circuits, in which at least a first cooling circuit comprises: a first cavity and a second cavity, the first and second cavities communicating with each other at a radially inner end and at a radially outer end of the blade.

Abstract: When cold and in the non-coated state, the aerodynamic profile is substantially identical to a nominal profile determined by the Cartesian coordinates X,Y, Zadim given in Table 1, in which the coordinate Zadim is the quotient D/H where D is the distance of the point under consideration from a first reference plane P0 situated at the base of the nominal profile, and H is the height of said profile measured from the first reference plane to a second reference plane P1. The measurements D and H are taken radially relative to the axis of the turbine, while the X coordinate is measured in the axial direction of the turbine.

Abstract: When cold and in the non-coated state, the aerodynamic profile is substantially identical to a nominal profile determined by the Cartesian coordinates X,Y, Zadim given in Table 1, in which the coordinate Zadim is the quotient D/H where D is the distance of the point under consideration from a first reference plane P0 situated at the base of the nominal profile, and H is the height of said profile measured from the first reference plane to a second reference plane P1. The measurements D and H are taken radially relative to the axis of the turbine, while the X coordinate is measured in the axial direction of the turbine.