An automated manufacturing system includes two simultaneous and independently operating toolheads accessing any location within the same work volume, with the exception of locations in proximity to each other. The system includes a bed platform connected with X and Y linear axes. A θ rotational axis rotates the bed and its linear axes as a unit. A first toolhead has a fixed position relative to the θ axis, and a second toolhead is coupled with a linear R axis parallel to the bed. The bed X and Y axes move the bed relative to the first toolhead, enabling the first toolhead to reach any portion of the bed. The R linear axis and θ rotational axis allow the second toolhead to move almost anywhere in a circular area that is always centered near the first toolhead. The system's kinematics ensure that it is impossible for the toolheads to collide.

The interference monitoring device comprises: an advanced position calculation unit that calculates an advanced position of the tool or the workpiece ahead by a predetermined period of time based on the program; an interference check unit that draws an estimated shape of a machined part of the workpiece after machining based on the advanced position, and makes the interference check based on the estimated shape of the machined part of the workpiece in the drawing; and an uncut region calculation unit that calculates an error between the estimated shape of the machined part of the workpiece in the drawing and an actual shape of the machined part of the workpiece after machining as an uncut region in the estimated shape of the machined part of the workpiece in the drawing. The interference check unit does not make the interference check in the uncut region.

A control device includes a machine learning device that learns a state of a spindle during normal machining without a collision of the spindle, and the machine learning device includes a state observation unit that observes spindle estimated load torque data indicating an estimated load torque value for the spindle and spindle acceleration data indicating an acceleration value of the spindle as state variables representing a current state of an environment and a learning unit that learns a correlation between the estimated load torque values for the spindle and the acceleration values of the spindle during the normal machining with use of the state variables.

An automated manufacturing system includes two simultaneous and independently operating toolheads accessing any location within the same work volume, with the exception of locations in proximity to each other. The system includes a bed platform connected with X and Y linear axes. A rotational axis rotates the bed and its linear axes as a unit. A first toolhead has a fixed position relative to the axis, and a second toolhead is coupled with a linear R axis parallel to the bed. The bed X and Y axes move the bed relative to the first toolhead, enabling the first toolhead to reach any portion of the bed. The R linear axis and rotational axis allow the second toolhead to move almost anywhere in a circular area that is always centered near the first toolhead. The system's kinematics ensure that it is impossible for the toolheads to collide.

The interference monitoring device comprises: an advanced position calculation unit that calculates an advanced position of the tool or the workpiece ahead by a predetermined period of time based on the program; an interference check unit that draws an estimated shape of a machined part of the workpiece after machining based on the advanced position, and makes the interference check based on the estimated shape of the machined part of the workpiece in the drawing; and an uncut region calculation unit that calculates an error between the estimated shape of the machined part of the workpiece in the drawing and an actual shape of the machined part of the workpiece after machining as an uncut region in the estimated shape of the machined part of the workpiece in the drawing. The interference check unit does not make the interference check in the uncut region.

A control device includes a machine learning device that learns a state of a spindle during normal machining without a collision of the spindle, and the machine learning device includes a state observation unit that observes spindle estimated load torque data indicating an estimated load torque value for the spindle and spindle acceleration data indicating an acceleration value of the spindle as state variables representing a current state of an environment and a learning unit that learns a correlation between the estimated load torque values for the spindle and the acceleration values of the spindle during the normal machining with use of the state variables.

An automated manufacturing system includes two simultaneous and independently operating toolheads accessing any location within the same work volume, with the exception of locations in proximity to each other. The system includes a bed platform connected with X and Y linear axes. A rotational axis rotates the bed and its linear axes as a unit. A first toolhead has a fixed position relative to the axis, and a second toolhead is coupled with a linear R axis parallel to the bed. The bed X and Y axes move the bed relative to the first toolhead, enabling the first toolhead to reach any portion of the bed. The R linear axis and rotational axis allow the second toolhead to move almost anywhere in a circular area that is always centered near the first toolhead. The system's kinematics ensure that it is impossible for the toolheads to collide.

An object of the present invention is to provide a numerical controller that performs interference check based on a look-ahead position before a feedrate varies by taking the variation in the feedrate into consideration and controls the change of an override safely. To achieve this object, the numerical controller is configured such that, when calculating a coordinate value of a look-ahead position which a movable part subjected to interference check reaches after a look-ahead period elapses from a current position at a current time, the numerical controller calculates the look-ahead position assuming that the movable part moves at a predetermined feedrate of which the upper limit is the maximum feedrate.

An inference device includes: a reception unit for receiving model information of a structure of a machine tool; a determination unit for determining, at the start of machining a polygon, an initial position of a center axis of the polygon, an initial phase of a rotary tool, and a position relationship between the center axis and a rotary axis of the rotary tool; a calculation unit for calculating a movement range of at least one of the rotary axis of the rotary tool and a rotary axis of workpiece based on the initial position of the center axis of the polygon, the initial phase of the rotary tool, and the position relationship between the center axis of the polygon and the rotary axis of the rotary tool; and an inference unit for inferring whether interference will occur in the machine tool, based on the model information and the movement range.

The invention relates to a method of controlling a gear cutting machine having at least one tool for cutting gears into a workpiece clamped in a workpiece mount, wherein the machine control carries out a three-dimensional collision monitoring during the gear cutting machining.