With huge amounts of data coming in from machines, some IoT vendors would make you think that’s all it takes. They can set up some alarms, alerts and flashing lights – and perhaps even send emails. From the surface, that may look interesting and bring some perceived value. The truth is, there is much more to it than simply aggregating data and when a threshold is hit, ringing the bell.

Machines on the production floor make items – that’s your life’s work and why you are in business. Each of these items have unique requirements in the production system.  Perhaps it’s a tolerance, or an oven temperature something must be cured at. Perhaps it’s the punch press dwell time that’s critical – too quick and you get cracking, too slow and you get parts that are too weak to perform.  Most times, production equipment will be built and set up to run multiple parts during the production day or week. Monday for 4 hours you will run 1000 of part “A” and then change to part “C” for the rest of the day. The monitoring of the machine and the data values generated will be specific and also need to be reported “In the context of the manufacturing part being produced”.  If the hydraulic pressure reads 2500 psi for part “A” this may be ok, but for part “C” we expect 1500 psi.  The data that gets collected and monitored must have the item process parameters set for the specific item being run.

Some IoT providers will say this is the operator’s responsibility to ‘adjust’ the machine to monitor the proper set-point limits.  In this case, we have asked the operator to perform a task that can best be handled programmatically.  Systems that can automatically set and reset limits, averages, set-points and other tag details based on the item being run will yield much more reliable results.  This is even more important as when this data is logged over time for deeper predictive analysis and trending, we need to have assurance the data is accurate and in context over time.

Bottom Line – raw data collected from machines, sensors and devices is not going to provide useful information unless that data is in the context of which item or part is being produced. With manufacturing IoT, machines run many parts and items – all with different metrics. The IoT data aggregation engine must be able to collect and manage this data within the constraints of the part specifications.

Stay tuned for part 3, where we’ll dive into Real Time Alerting for manufacturing environments.

Learn more at www.Godlan.com