Machine tool and machine learning device

1. A machine tool, comprising:

a plurality of control circuits for moving a spindle including a tool relative to a workpiece by controlling at least one axis on the basis of a machining program and machining the workpiece; and

a machine learning device which includes a processor configured to predict occurrence or non-occurrence of a collision between

(i) the spindle or the tool, and

(ii) the workpiece, a jig holding the workpiece, or a constituent element of the machine tool,

said collision caused by an operation for the axis through manual operation after interruption of the machining, wherein

the machine learning device further includes a storage device which stores a learning model constructed by supervised learning using

(i) input data based on information related to the manual operation, and

(ii) teacher data based on information related to the occurrence or non-occurrence of the collision of the spindle or the tool at the time of the manual operation, and

the processor is further configured to

acquire the information related to the manual operation from log data recorded at the time of the manual operation in the machine tool,

create the input data on the basis of the information related to the manual operation acquired from the log data, wherein the input data include (i) first input data in a case where a spindle collision occurred due to an improper manual operation by an operator, and (ii) second input data in a case where no spindle collision occurred due to an improper manual operation by an operator, and

predict, using the learning model, the occurrence or non-occurrence of the collision of the spindle or the tool from the input data created based on the information related to the manual operation.

2. The machine tool according to claim 1 , further comprising:

an alarm unit configured to output an alarm on the basis of a result of the prediction of the occurrence or non-occurrence of the collision of the spindle or the tool from the machine learning device.

3. The machine tool according to claim 2 , wherein the machine learning device is further configured to share the learning model stored in the storage device with at least one different machine tool.

4. A machine learning device for, in a machine tool which moves a spindle including a tool relative to a workpiece by controlling at least one axis on the basis of a machining program and machines the workpiece, learning occurrence or non-occurrence of a collision between (i) the spindle or the tool and (ii) the workpiece, a jig holding the workpiece, or a constituent element of the machine tool, said collision caused by an operation for the axis through manual operation after interruption of the machining, the machine learning device comprising:

a processor configured to

acquire information related to the manual operation from log data recorded when the machining and the manual operation are performed in the machine tool,

create input data on the basis of the information related to the manual operation acquired from the log data, wherein the input data include (i) first input data in a case where a spindle collision occurred due to an improper manual operation by an operator, and (ii) second input data in a case where no spindle collision occurred due to an improper manual operation by an operator,

acquire information related to the occurrence or non-occurrence of the collision of the spindle or the tool at the time of the manual operation from the log data recorded when the machining and the manual operation are performed in the machine tool,

create teacher data on the basis of the information related to the occurrence or non-occurrence of the collision of the spindle or the tool acquired from the log data,

perform supervised learning using the input data created based on the information related to the manual operation and the teacher data created based on the information related to the information related to the occurrence or non-occurrence of the collision of the spindle or the tool, and construct a learning model; and

a storage device which stores the learning model.

5. A machine learning device for, in a machine tool which moves a spindle including a tool relative to a workpiece by controlling at least one axis on the basis of a machining program and machines the workpiece, predicting occurrence or non-occurrence of a collision between (i) the spindle or the tool and (ii) the workpiece, a jig holding the workpiece, or a constituent element of the machine tool, said collision caused by an operation for the axis through manual operation after interruption of the machining, the machine learning device comprising:

a storage device which stores a learning model constructed by supervised learning using

(i) input data based on information related to the manual operation, and

(ii) teacher data based on information related to the occurrence or non-occurrence of the collision of the spindle or the tool at the time of the manual operation;

a processor configured to

acquire the information related to the manual operation from log data recorded at the time of the manual operation in the machine tool,

create the input data on the basis of the information related to the manual operation acquired from the log data, wherein the input data include (i) first input data in a case where a spindle collision occurred due to an improper manual operation by an operator, and (ii) second input data in a case where no spindle collision occurred due to an improper manual operation by an operator, and

predict, using the learning model, the occurrence or non-occurrence of the collision of the spindle or the tool from the input data created based on the information related to the manual operation.

6. The machine learning device according to claim 4 , wherein the machine learning device is further configured to share the learning model stored in the storage device with at least one different machine learning device.