For as long as there have been factories, manufacturers have sought ways to improve productivity. That quest now defines entire professional disciplines. And it has certainly resulted in steady, even remarkable, improvements in productivity – generally speaking. It’s also generated countless technologies, methodologies, protocols and philosophies that directly or indirectly seek to drive better manufacturing results, achieving lower costs, higher revenues, or both. And disparate data sources and inconsistent metrics can compound this complexity, resulting in misleading and irreconcilable findings.
Art, Science, and Intuition
Like every modern manufacturer, the pressure you face is not only to improve – and improve continuously – but also to prove you’re doing it. “Improvement” per se can be directionally clear across the various metrics by which it’s tracked, but effective comparison across different metrics is difficult. Establishing the “whole truth” about productivity is difficult enough without the complicating factor of disparate methods and metrics. As a result, over the decades, assessing and managing productivity in your factory has taken on aspects of “art”. The intuition of manufacturing professionals becomes a reporting filter – even as we’ve saturated the enterprise with more “science” in all the forms discussed above. When that happens, the truth of a situation may lie in the subjective eye of the beholder.
The Inherent Complexity of OEE
Despite its longstanding utility, Overall Equipment Readiness (OEE), the most widely accepted measure of how well a manufacturing operation is performing, can also add further complexity to the production improvement ecosystem. Multiple equations for calculating OEE exist, each considered valid in the environment where deployed. One simplest valid OEE calculation, (Good Count x Ideal Cycle Time)/Planned Production Time), doesn’t factor in certain loss- related variables, while another (Availability x Performance x Quality) is centered on those very factors.
From Six Sigma to Lean, SCADA to MES, enterprises continue to deploy an ever-increasing variety of techniques to better understand, control, and optimize how their factories run. Such a wide array of approaches necessarily drives staggering complexity. (MESA, the Manufacturing Enterprise Solutions Association, was founded to deal with that complexity – and that was 30 years ago.) This complexity means disparate data sources and measurement standards, even in single factories where a casual observer would expect some reasonable degree of harmonization.
Further, the things you measure to assess OEE are themselves different in kind, and so traditionally described according to units that are not immediately comparable. The measure of planned and unplanned downtime, and changeover, for example, are normally measured in hours. Production speed, in units per hour. Minor stops, and scrap- always in absolute units. The implications of these metrics are not commonly oriented for a single, comparable view of reality.
Inconsistent Data Sources
Meanwhile, mission-critical data is collected and resides in different places in the manufacturing ecosystem. It’s aggregation presents challenges as well. These data sources may be not only discrete from one another, but different in kind: ERP systems, your MES, PLCs, IoT data, SCADA systems. And of course there are often homegrown point solutions, the intricacies and attributes of which may be fully understood only by a limited team.
The problem with percentages
To identify production bottlenecks and opportunities, discern cause and effect, and take most effective actions, organizations often default to percentage-based reporting, wich itself cab be inherently misleading. Because the inputs to reporting arrive from so many varied sources, the picture that emerges may be misaligned; less coherent, less accurate, and less effectively predictive of the best action to take. A large percentage improvement in a less-than-critical problem may be less optimal choice than a smaller improvement in a more critical step. Furthermore, such reporting tends to be retrospective; a view of what was happening at the time a given measurement took place, further delayed while processed through the many filters and computations we’ve discussed.
Making your case
Data is everywhere. But where’s the truth? Given the daunting complexity in which competing systems, multiple variables, and disparate data sources result, it can be tough to discern. Given the challenges inherent to comparing percentage-based reporting derived from heterogenous data types across multiple sources, how can you, as an operations manager, clearly discern and prioritize the best, highest-impact next step? And how do you make your case in language that’s mutually intelligible to all your constituents?
One language, one view
With the introduction of ThingWorx, Digital Performance Management, PTC has shifted the paradigm. DPM was conceived and engineered to realize a single, common vocabulary for assessing, prioritizing and addressing the issues that will have the biggest impact on your manufacturing productivity. All the complexity that defines data flowing off the modern production line is resolved and rationalized to a unified metric – production hours – with an unambiguous, commonly- understood meaning. The truth – and the path to measurable, demonstrable production improvement – emerge more clearly that ever before.
For more information about ThingWorx, Digital Performance Management contact us at: firstname.lastname@example.org