top of page
SM Real-Time Performance Dashboard
The "SM Real-Time Machine Performance & Productivity Dashboard" is a real-time data analytics dashboard calculates analytics on the data streams from the cloud. The dashboard streams the data from two sensors power and accelerometer. It is capable of synchronously streaming and analyzing the data at a sampling rate of 10kHz. It has a displays of metrics such as signal visualization, machine state, productivity metrics, quality metrics and unique signal features.
Featured Capability
-
Continuous health monitoring and productivity of the machine.
-
A time lag of less than 5 seconds between the actual operation and its display on the dashboard.
-
Visualization of operations using time portraits, FFT and spectrogram to detect anomalies.
-
Real-time prediction of surface roughness with an accuracy of 90% using a tuned algorithm.
-
Detection of abnormal parts produced.
Machine State
The machine state LED will display the current state of the machine as to whether it is stopped or is running in idle, normal or an abnormal operation mode. It is classified based on the signal from the power sensor of the machine. The explanation of different states is given below: -
Stop - The machine is either switched off or the spindle is not rotating.
Idle - The machine spindle is rotating idly, and the workpiece has contacted the grinder. It refers to the time when the machine is on, but no work is being done on the workpiece.
Normal Operation - It refers to the conditions when the workpiece is machined normally without any problems.
Abnormal Operation - It refers to the conditions during machining when the power requirements are higher than observed normally. This might be due to tool wear, broken tool tip or other faults occurring during machining.
Quality Metrics
The quality metrics keeps track of the quality of the product and can alert the operator in case of any anomality occurring during the machining. It helps in flagging anomalous products which can then be removed from further processing and prevent unnecessary losses down the line. The quality metrics are:-
Surface Roughness - By using a tuned random forest algorithm, the surface roughness for each part machined is calculated from the sensor data. The predicted surface roughness value is a good indication as to whether the part is machined correctly or not.
Anomalous Parts - The number of parts manufactured during which an anomaly was detected from the sensor data.
Good Parts - The number of parts which are manufactured without a known anomaly and are assumed to be good.
Grinding Burns - Number of parts which have a grinding burn during machining.
Productivity Metrics
The productivity metrics keeps track of the overall productivity of the machine. It lets the operator know how many parts are produced, idle time, run time and the down time of the machine. The productivity metrics are: -
Part Count - Total number of parts produced by the machine.
Run Time - The total time the machine is running.
Idle Time - The time during which the machine is consuming power but there is no interaction between the workpiece and the tool.
Down Time - The time during which the machine is stopped and is not working. It could be due to changing of the workpiece or for performing maintenance.
bottom of page