Sometimes this feature itself may pay the whole system off. Check out the example in our case study. This works equally well for both, palletized systems and index tables.
Reject chart shows number of rejects for a specific pallet (fixture). This can be shown for total, or per reject code (operation).
By selecting a reject code the chart shows number of times a pallet was rejected on a specific operation – station.
This will never replace a QC department, but is a great help to have near a machine – especially when something goes wrong.
This is a ‘live’ WonderWare screen shot with a sketch of an unload station. Green squares represent active sensors (proxy ON). Gray ones are off. Blue triangle is an active air valve representing direction of the force. Above and near the sensors and valves are wire numbers. Active sequence steps are lit yellow, note ‘Done’ in the sequence list.
This certainly beats any ‘message based’ diagnostic system and is way simpler. Suppose the sensor 5000 (top right) is off and the valve 8020 is on. What does it mean? Indeed, the picture is worth thousand diagnostic messages.
It is even possible to ‘move’ objects on the screen to achieve full animation of the machine, though one can argue usefulness of that bearing in mind that the PC is actually located near the machine.
Theoretically it is possible to stop a line (machine) before it starts making rejects. Based on several statistical rules we can detect various machine drifts and raise an alarm although all the parts that a machine is making are still good.
For this we would normally use a readily available SCADA package with the SPC module, like WonderWare with SPC Pro.
SPC alarm lists bunch of rules we can use to generate an alarm and stop the machine.