What happens when MOM systems collect and distribute inaccurate data?

Much of the decision-making in plant operations (pre-shift and in-shift) requires real-time raw data collections to be accurate in terms of timing, complexity of the process and context. If any of this data is delayed, collected out of order or contains errors, the aggregated data may lead to inaccurate decisions for scheduling, resource allocation, product quality, process analysis and performance and planning metrics.

Often production order reporting data is collected manually with little to no enforcement of timely recording. And in general, even when automated or manual data collections are timely and used accurately in the applied operation-level context, the decision accuracy can be lost. How and why?

Operations metrics, events and comparative analysis typically use raw data from 15 to 50 standalone plant systems (Scheduler, Production Track/Trace, Maintenance, Quality, Warehouse, Batch, PLCs, Historians, etc.) all with different database structures. In order to allow for intelligence-based decision-making, each of these disparate data sources must be aggregated into one decision data set. This requires manual analysis and calculation, which is prone to errors.

The above issues affect many decision making metrics or KPIs including: OEE metrics, ERP cycle time standards comparisons, plant WIP inventory consumptions compared to finished goods counts, warehouse counts used for logistics scheduling and full-load shipping, NPI time-to-volume and time-to-margin, etc.

So, to come full circle to the question posed: What happens when MOM systems collect and distribute inaccurate data? It causes bad decision-making!

How do you avoid these issues? Implement a MOM solution that uses independently time-encoded streams (both for storage and retrieval) of sensor data, contexts and events. This allows the system to infer, combine and enrich information based on a model and a context of execution. A MOM solution with this technology ensures that while data is collected in real-time, analysis and calculations remain highly accurate and predictive.