A machine tool diagnosis method is operated in conjunction with a machine tool diagnosis system, which uses a convolutional neural network (CNN) model to analyze machining signals to generate target information including processing results and diagnosis opinions, and graphically displays a machining path of a target machining object on a display interface according to the processing results, where the display interface will further display the diagnosis opinions corresponding to the machining path.

Provided is a dicing apparatus and a method for controlling a dicing apparatus that can discover undivided defects in a die attach film.

A dicing apparatus 10 that performs dicing of a street of a workpiece W attached to a dicing tape 9 via a die attach film (DAF 7) and cuts the workpiece W and the die attach film along the street includes: a cross-sectional profile acquisition unit (a white interference microscope 24 and a processing unit 64) that acquires a cross-sectional profile of a machined groove 19 formed through the dicing; a cut amount detection unit 72 that detects a cut amount ?d into the die attach film obtained through the dicing on the basis of the cross-sectional profile acquired by the cross-sectional profile acquisition unit; and a division determination unit 74 that determines whether or not the die attach film is in an undivided state on the basis of the cut amount detected by the cut amount detection unit 72.

A processing system having a function for appropriately maintaining processing accuracy, without depending on a degree of abrasion of a tool. An unused tool used in a machine tool is captured by using a camera mounted on a robot, so as to obtain a reference image of the tool. Next, an image of the tool is captured after a predetermined number of processing operations carried out. An average amount of tool abrasion per one processing operation is calculated based on the two images, so as to estimate a current amount of tool abrasion. Based on the estimation result and a previously input limit amount of abrasion, the number of remaining possible processing operations of the tool is calculated and output.

A machining dimension prediction apparatus includes a trend acquirer that acquires, for each workpiece, trend information indicating a trend of a state of a machining tool during a machining period of machining performed by the machining tool, a feature calculator that calculates, based on the trend information, a feature using the trend of the state in each of sections included in the machining period, a measurement value acquirer that acquires a measurement value of a dimension of each workpiece after being machined, a section specifier that specifies, as a specific section of the sections, a section including a calculated feature having a greatest degree of relevance to the measurement value, and a predictor that predicts, when a new-target workpiece is machined, a dimension of the new-target workpiece after being machined based on the feature calculated using the trend of the state in the specific section.

A processing system having a function for appropriately maintaining processing accuracy, without depending on a degree of abrasion of a tool. An unused tool used in a machine tool is captured by using a camera mounted on a robot, so as to obtain a reference image of the tool. Next, an image of the tool is captured after a predetermined number of processing operations carried out. An average amount of tool abrasion per one processing operation is calculated based on the two images, so as to estimate a current amount of tool abrasion. Based on the estimation result and a previously input limit amount of abrasion, the number of remaining possible processing operations of the tool is calculated and output.

A machining dimension prediction apparatus includes a trend acquirer that acquires, for each workpiece, trend information indicating a trend of a state of a machining tool during a machining period of machining performed by the machining tool, a feature calculator that calculates, based on the trend information, a feature using the trend of the state in each of sections included in the machining period, a measurement value acquirer that acquires a measurement value of a dimension of each workpiece after being machined, a section specifier that specifies, as a specific section of the sections, a section including a calculated feature having a greatest degree of relevance to the measurement value, and a predictor that predicts, when a new-target workpiece is machined, a dimension of the new-target workpiece after being machined based on the feature calculated using the trend of the state in the specific section.

The wear status of a micro-endmill tool may be inferred by monitoring the chip production rate of the tool in operation. Chips may be extracted from a work area, captured on an adhesive surface, imaged, and counted to determine the chip production rate. When the rate of chip production falls, the feed rate of the micro-endmill may be modified to a level suitable for the current state of tool wear. In this manner, costly and inconvenient work stoppages to evaluate the wear status of a tool are eliminated.

A cutting blade partially covered with one or more layers covering a part of the blade, with one of the layers being a wear layer which has a different visual appearance than the underlying material of the blade, and wears away more quickly than the core material when the blade is used, so that the underlying material of the blade is uncovered as soon as this wear layer is completely removed in partial areas, providing a clear visual signal that it needs to be replaced.

Maximum values of actual machining allowances of a workpiece in respective cutting flutes are averaged by control amounts of respective cutting flutes calculated by an arithmetic unit. The arithmetic unit converts the calculated control amounts into an amount in each axis direction and subtracts the value from a feed axis command value in an NC program stored in a storage unit. Then, a numeral value control unit performs machining by controlling respective feed axes based on the subtracted feed axis command value.

The wear status of a micro-endmill tool may be inferred by monitoring the chip production rate of the tool in operation. Chips may be extracted from a work area, captured on an adhesive surface, imaged, and counted to determine the chip production rate. When the rate of chip production falls, the feed rate of the micro-endmill may be modified to a level suitable for the current state of tool wear. In this manner, costly and inconvenient work stoppages to evaluate the wear status of a tool are eliminated.