How Advanced Analytics Enables OEE Improvement

Summary

Overall equipment effectiveness (OEE) is a standardized and internationally recognized performance metric used to represent the effectiveness of a line, machine, or process. Expressed as a percentage, it quantifies the proportion of planned value-adding production time, and how much time and resources are being wasted due to unplanned availability losses, performance issues and quality losses.

Popularized in the 1980s, OEE has evolved from a rough calculation scrawled on a shopfloor chalkboard to a critical key performance indicator (KPI), with an entire software market dedicated to digitally monitoring and enhancing its score (Figure 1).

Limitations of OEE

With an OEE of 85 percent often touted as “world-class” or the “gold standard,” this might give the misguided impression that it is the sole marker of a successful and productive operation. In reality, the situation is much more complex, and fixating on this single metric—rather than using it as a component of a broader framework—can be detrimental to a company’s overall objectives and bottom line.

When OEE is treated as the singular target, there arises a potential incentive for workers to manipulate inputs to achieve their targets. For example, tweaking the ideal production rate or misclassifying production stops as planned events are two common methods for artificially inflating a machine’s OEE without any meaningful increase in production or profit. This is also problematic because it decreases the quality of data collection, which ultimately causes inaccurate or misguided data-based decision-making.

OEE is focused on the local optimization of an individual unit. However, the end goal should be optimizing the entire production system. To effectively achieve this, the context of wider topics—such as market prices, energy consumption and sustainability goals—must be considered.

Furthermore, OEE is an effective measurement for assessing the current state of operations and potential for efficiency gains, but monitoring OEE in isolation does not produce improvements. For example, knowing a machine failed because of a certain fault, or quantifying the number of microstops that occurred in the past week, is of limited value without the ability to pinpoint root causes, predict future events and take corrective measures to reduce subsequent occurrences.

According to IDC Insight’s 2022 Worldwide IT/OT Convergence Survey, the average cost of downtime across industries is $200 k/hr, so the benefits of even small incremental improvements are significant. When continuous improvement efforts can be accelerated or enhanced with digital solutions, like advanced analytics software, realizing return on investment occurs quickly.

Breaking down data silos with advanced analytics

As with any data-driven decision-making, the value of the outcome largely depends on data availability, accessibility and quality. Calculating OEE metrics tends to require very limited data, but fully addressing the underlying issues often necessitates more elaborate datasets.

When digitally mature companies implement OEE improvement strategies, data availability is rarely the limiting factor. Instead, the primary barrier is typically data accessibility, with relevant information spread throughout many different data sources such as a process historian, manufacturing execution system (MES), laboratory information management system (LIMS), or other similar systems. It is also usually in a raw format that requires significant data cleansing and contextualization to provide meaning.

This was the finding of one multinational consumer goods company looking to address the issue of frequent micro-stops in its process. Information about micro-stop events was contained within the manufacturer’s MES, and configuration settings were stored in the process historian. With this data siloed in different sources, it was previously impossible to find correlations between settings and the frequency of micro-stops without complex and time-consuming data wrangling.

However, by deploying an advanced analytics solution, the company was able to access multiple data sources from one central location, empowering users to seamlessly combine and interrogate data regardless of source. This enabled a simple correlation analysis that provided users with rapid diagnostics to identify optimal configuration settings, resulting in significant performance improvements.

The other principle of data accessibility is providing all users with access to the information they need when they need it. In the past, the manual collection of data meant OEE could only be reported on days or weeks after the fact. This activity generated some interesting insights for management, but operations staff often lacked the ability to make proactive improvements. Even now, with the majority of OEE monitoring systems collecting and calculating metrics automatically, the results are often presented in a way that is more aligned with the long-term reporting method of the past, instead of near-real-time monitoring.

It is important to carefully consider the information an operations team and supporting functions need, and to provide them with relevant, real-time and auto-updating dashboards that reflect this. For example, a world-renowned pharmaceutical company developed multiple dashboards that were designed with its operations staff in mind. One of the most impactful dashboards provided visualizations of the current duration of multi-step changeovers.

Because changeovers are a mandatory process step, they are usually categorized as a planned stop reason that does not contribute toward availability loss. However, the company was experiencing a high variability in the duration of changeovers, which was limiting the amount of finished product. By leveraging an advanced analytics solution, the company was able to accurately quantify the lowest repeatable time for each stage of the changeover and use a simple formula to split the changeover into components of planned and unplanned downtime. A custom-built dashboard then highlighted all changeover durations that were excessive as events requiring investigation (Figure 2).

This additional granularity was presented directly to the operations staff in the form of a traffic-colored process flow, providing the opportunity to immediately investigate and resolve delays. These proactive investigations significantly reduced turnaround time variability, which in turn increased time spent in production.

Innovate to add value

Calculating and monitoring OEE remains a highly valuable metric for measuring equipment effectiveness. However, it is not sufficient to mindlessly collect and monitor the data. It must be translated and converted into meaningful, actionable information for various end users to provide business value.

By fostering improvement culture and empowering workers with collaborative advanced data analytics technologies, organizations can reach new levels of efficiency needed to remain competitive in today’s fast-paced manufacturing market.

_{All figures courtesy of Seeq
This feature originally appeared in the December 2023 issue of InTech digital magazine.}