A method for training a neural network to derive, from an image of a camera mounted on a robot, a movement vector to insert an object into an insertion. The method includes, for a plurality of positions in which the object held by the robot touches a plane in which the insertion is located controlling the robot to move to the position, taking a camera image by the camera and labelling the camera image with a movement vector between the position and the insertion in the plane and training the neural network using the labelled camera images.

A machine learning device acquires from a numerical controller information relating to machining when the machining is performed, and further acquires an actual delay time due to servo control and due to machine movement which are caused in the machining when the machining is performed. Then, the device performs supervised learning using the acquired machining-related information as input data, and using the acquired actual delay time due to servo control and due to machine movement as supervised data, and constructs a learning model, thereby predicting the machine delay time caused in a machine with high precision.

To provide a wire electric discharge machine which can appropriately perform thermal displacement correction of upper/lower guides even when the installation environment changes. Provided are a storage unit that stores temperatures of machine elements as temperature data, and a rendering unit that digitizes the installation environment and renders as environmental data. Additionally provided are a position command unit that commands a relative position of the upper/lower guides; and a relational expression calculation unit that sets the temperature data environmental data as input data, sets the relative position as training data, and calculates the relational expression by way of machine learning. Further provided are a relational expression decision unit that calculates a correction amount by substituting the temperature of the machine element into this relational expression, and in the case of error between the relative position of the upper/lower guides based on this correction amount and the relative position commanded by the position command unit being small, decides this relational expression as a formal relational expression; and a correction execution unit that performs correction on the relative position of the upper/lower guides using this relational expression.

The embodiments of the present invention provide a dynamic production scheduling method, apparatus and electronic device based on deep reinforcement learning, which relate to the technical field of Industrial Internet of Things, and can reduce the overall processing time of jobs on the basis of not exceeding the processing capacity of production device. The embodiments of the present invention includes: acquiring static characteristics, dynamic characteristics of each of jobs and system dynamic characteristics, inputting the static characteristics, dynamic characteristics of each of jobs to be scheduled and system dynamic characteristics into a scheduling model to obtain a job execution sequence or batch execution sequence of the jobs in each production stage, wherein, the static characteristics of the job include an amount of tasks and time required for completion, the dynamic characteristics of the job include reception moment, and the system dynamic characteristics include a remaining amount of tasks that can be performed by the device in each production stage. The scheduling model is a model obtained after training a first actor network based on static characteristics and dynamic characteristics of a sample job, system dynamic characteristics, and a first critic network.

A method for improving a cycle time of a process of a product is provided. The method includes: collecting process profile data from a plurality of tool groups running the process, and calculating values of a plurality of key-performance-indicators (KPIs) of each tool group including calculating a standard deviation of an output of a stage of a bottleneck tool group of the tool groups; feeding the values of the KPIs and a work-in-progress (WIP) of each tool group into a neural network model in order to output an impact on the WIP for each KPI of each tool group by the neural network model; selecting a set of major KPIs of each tool group from the KPIs according to the impact of each tool group; and controlling the tool groups according to the impact of the set of major KPIs of each tool group in order to reduce a total WIP.

A method for improving a cycle time of a process of a product is provided. The method includes: collecting process profile data from a plurality of tool groups running the process, and calculating values of a plurality of key-performance-indicators (KPIs) of each tool group including calculating a standard deviation of an output of a stage of a bottleneck tool group of the tool groups; feeding the values of the KPIs and a work-in-progress (WIP) of each tool group into a neural network model in order to output an impact on the WIP for each KPI of each tool group by the neural network model; selecting a set of major KPIs of each tool group from the KPIs according to the impact of each tool group; and controlling the tool groups according to the impact of the set of major KPIs of each tool group in order to reduce a total WIP.

The present disclosure relates to an apparatus for printing three-dimensional, 3D, objects by a 3D printer. The apparatus comprises interface circuitry configured to receive a G-code file comprising instructions for 3D printing of one or more 3D objects. The apparatus further comprises processing circuitry configured to classify a 3D object represented by the G-code file, and restrict 3D printing of the 3D object if the 3D object is classified as at least a part of a dangerous/illegal object.

For optical quality control, a product image of a production installation captured by an image capture device for given intrinsic and extrinsic parameters is used and digital twin data of a digital twin of the production installation are used, to render a synthetic simulation image based on the digital twin data, wherein the rendered synthetic simulation image is based on the same intrinsic and extrinsic parameters as during product image capture, to transfer the product image from a real domain into an artificial domain by a trained domain adaptation and in the process to generate a synthetic product image from the product image with domain transfer parameters obtained by the training, to compare the synthetic product image with the synthetic simulation image by a comparison operator, and to output a comparison result which qualitatively assesses the product.

Provided is a control apparatus comprising a control unit configured to control a control target by a control model machine-learned so as to output an operation amount of the control target according to a state of equipment provided with the control target; a simulation unit configured to simulate, by using a simulation model, the state of the equipment in a case where the operation amount, which is output by the control model, is given to the control target; and a stop unit configured to stop control of the control target by the control model, based on a simulation result.

A tactile sensing system of a robot may include: a plurality of piezoelectric elements disposed at an object, and including a transmission (TX) piezoelectric element and a reception (RX) piezoelectric element; and at least one processor configured to: control the TX piezoelectric element to generate an acoustic wave having a chirp spread spectrum (CSS) at every preset time interval, along a surface of the object; receive, via the RX piezoelectric element, an acoustic wave signal corresponding to the generated acoustic wave; select frequency bands from a plurality of frequency bands of the acoustic wave signal; and estimate a location of a touch input on the surface of the object by inputting the acoustic wave signal of the selected frequency bands into a neural network configured to provide a touch prediction score for each of a plurality of predetermined locations on the surface of the object.