The Importance of Data Analytics in Manufacturing for Boosting Your Plant's Operating Efficiency

June 26, 2018

The manufacturing world has embraced big data. Large corporations and individual facilities alike have completed and reflected on studies based on the analysis of large amounts of data they have gathered from their products, processes, and supply chains.

From these big data studies, manufacturers have learned two primary lessons:

  1. Data analytics for manufacturing must be tied to essential business questions in order to be valuable.
  2. When data analytics are properly focused on these business questions, they can generate great value.

While analytics can be used to gain valuable insight into many aspects of large companies, one of the areas where plants stand the most to gain is in improving operational efficiency. As a Plant Operations Manager, you can maximize asset reliability and minimize system downtime by employing analytical tools. This can increase yield, improve product quality, and raise profits.

Boosting Manufacturing Plant Operating Efficiency with Data Analytics

Within a production facility, the most reliable predictors of profitability are equipment uptime (which translates to product yield) and product quality. The product yield of a plant within a given time frame, multiplied by the plant’s defect rate percentage within that same time frame, produces a total number of defective products. This defects number, subtracted from the original product yield number, returns the total yield of your plant—again, within the same specified time frame.

[Product Yield] x [Defect%] = [Total Defective Products]

[Product Yield] – [Total Defective Products] = [Total Yield]

If you’re reading this, you already know that the income from sales of total product yield, less the costs of operating your plant, determines your profitability. And, your profitability determines your potential for growth—and for future plant improvements (and maybe even your bonus). As total yield is directly related to product yield and the defect rate, profitability is directly affected by improvements in either yield or in product quality.

Analyzing the Factors Affecting Product Yield and Quality

The analysis of large amounts of raw asset data can uncover insights into ways to increase yield, often via improved asset performance management. For instance, consider performing a comparative uptime analysis of similar pieces of the same type of equipment within each plant under a large corporation’s umbrella. This can determine factors contributing to, and indicators of, unexpected downtime, allowing a Plant Operations Manager to avoid unplanned downtime with timely maintenance or repair. Deterring downtime obviously increases product yield by keeping your equipment up and running.

Understanding the factors which cause product failure allows you to make the maintenance or process adjustments you need to reduce or prevent lost product in the future.

data analytics for manufacturingLikewise, comparing equipment factors vs. product defect reports can help manufacturers understand the reasons behind a bad batch of product. Without this insight, millions of dollars could be lost in botched product without a single lesson learned. Understanding the factors which cause product failure allows you to make the maintenance or process adjustments you need to reduce or prevent lost product in the future.

The more you learn about your equipment, the better you’ll be able to keep your equipment running. The better you know what conditions negatively impact your product, the closer you’ll come to a reliable, efficient process. Your grandfather probably could’ve told you that. So why the big deal about big data?

Analyzing Manufacturing Plant Asset Data—and Implementing Lessons Learned

Your grandfather, large as he may loom in your memory, was only one single person. The human mind, while a highly capable computational engine, has hard limits on memory and processing speed. In our current, tech-driven world, we can now do better.

Modern big data analytics are exactly what they sound like: big. Information from hundreds or thousands of similar assets can be collected, compiled, and compared. Not only can we now digitally collect and store previously unheard-of volumes of information, we also make sense of that collected data as well. A well-developed algorithm can sift through terabytes of sensor data to track dozens of variables that affect your equipment, note how the variables affect each other, and highlight data anomalies for extra scrutiny.

Your assets are complex systems, built up out of smaller, yet themselves complex, systems, and are affected by the conditions present in an even more complex system: their external environment. To factor the relationships between ambient relative humidity, belt replacement schedules, and product defect rate, for example—let alone the dozens of other factors influencing your yield and quality—is only possible with the aid of modern information technology.

Implementing Lessons Learned from Analytics

The greatest library in the universe is useless if nobody checks out a book. All the data you can collect on every piece of equipment in the world is worthless if it isn’t analyzed. This analysis, likewise, is meaningless unless it produces actionable steps you can take to improve your plant’s profit margin.

In order to glean useful, actionable insights from your analytics, you first need to be asking the right questions. Here’s a study guide to a useful application of big data:

  • Depending on the specifics of your plant, focus your inquiry on either equipment uptime or product quality.
  • Run algorithms to determine the causes behind your equipment downtime and defect rates.
  • Prioritize the greatest sources of loss.
  • Determine what corrective actions you can and should take.
  • Implement those corrective actions.

As you make changes to your processes and equipment maintenance procedures, be sure to continue to collect data to determine the effect of the changes you’ve put into place. This will not only show you how well your changes are working, it will also inform your next steps.

Timely, Relevant Data Analysis Improves Plant Process Efficiency

It doesn’t take just heat and pressure to make diamonds—it also takes coal.

No amount of precise, powerful algorithmic calculations will produce timely, relevant insights unless these calculations are performed on timely, relevant data.

No amount of precise, powerful algorithmic calculations will produce timely, relevant insights unless these calculations are performed on timely, relevant data. Equipment built 20 years ago is not the same as equipment built last month. Environmental data from Qatar is not relevant to conditions in Alaska. The more general the information is, the more general your insights will be as well.

The need for timely, relevant data is particularly pressing when developing maintenance schedules. Maintenance schedules based on average data will calculate and predict average failure intervals—not actual failure intervals. While average failure intervals will allow you to come closer to preventing unexpected equipment downtime than no failure intervals will, they won’t allow you to prevent downtime entirely.

Real-Time Data Collection Allows Conditional Maintenance

In order to predict and prevent equipment downtime in real time, it’s necessary to collect real-time data on equipment conditions. Analyze the condition of your assets leading up to failure. This will provide you with insight into which conditions predict failures. Then, as you monitor your assets in real time, when pre-failure conditions arise, you will be able to appropriately respond to prevent downtime.

This may be an inconsequential issue for small or enclosed facilities which are well-monitored by personnel and automatic sensors. But for large facilities, spread over acres, timely collection of asset data can be a significant blind spot. This can be a serious problem for larger plants as downtime of large assets in chemical plants, oil refineries, and power plants can cost hundreds of thousands of dollar per hour in lost product, if not more.

Mobile Data Collection Improves on Paper Checklists

Still, much of the relevant data that Operations Managers need to collect to perform their analysis can’t be instrumented, and is recorded by mobile workers on paper checklists. For instance, evidence of leaking, excessive vibration, or irregular noises must be noticed in person.

By the time predictors of failure are noticed, the asset may have already broken down and caused downtime.

The checklists these issues are recorded on may not be viewed for days, or even weeks. By the time predictors of failure are noticed, the asset may have already broken down and caused downtime. Just as many manufacturing operations are moving to big data analytics to improve operating efficiency, so are many large plants and facilities embracing mobile asset data collection technology.

When conditional information on large, important assets is entered into a tablet or smartphone, it can be registered and analyzed immediately by your plant’s central processing system. As a result, you’ll be able to detect and prevent impending equipment failure.

In addition to your regular preventative maintenance schedule, you can implement a conditional maintenance schedule. When your Operator notices a leak, they’ll notify you immediately via their mobile device. Then you can send a team to fix it before your pump runs out of lubrication and stalls. You’ll be able to catch problems before they happen, and stop them from ever coming to pass.

Modern information technology, including the insights gained from focused analysis of big data and the real-time reporting ability of mobile technology, has the potential to dramatically impact the bottom lines of manufacturing plants and facilities. Learning the conditional predictors of lost product or equipment failure, and being able to detect and intercede before they result in equipment downtime or substandard product, allows you to improve your product yield, your product quality, and, ultimately, your profit margin.

At GoPlant, we’re working hard to bring your plant the modern tools it needs to stay profitable in the 21st century. Our mobile, asset-centric data collection platform enables real-time monitoring of your most important equipment. To see GoPlant in action, request a no obligation demo. Or, get in touch with our team today.

Image courtesy Manine99