Why Use Predictive Analytics in Manufacturing Quality Management?
Get realistic and attainable results when you look more closely at the data you’re already collecting.
Predictive analytics, or the analysis of all incoming data to identify problems in advance, is a fairly common topic in manufacturing boardrooms and management meetings. But the conversation often revolves around trying to understand the purpose or value of spending money to implement analytics in a manufacturing environment.
Respondents to a study by Honeywell said they believe data can:
- Enable well-informed decisions in real time (63%)
- Limit waste (57%)
- Predict the risk of downtime (56%)
Unfortunately, in the more than 30% of respondents who don’t have predictive data analytics in place and don’t plan to add it in the next year, many believe their companies can grow without it, that they already have what they need, or the benefits are “overrated.”
Are $1 million in sustainable annual savings and a drastic plant turnaround from worst to top performer in six months overrated?
You tell me. These realistic and attainable results were achieved by one of our customers who properly implemented predictive analytics. Let’s break down the thought process.
1. Gather the Right Quality Data
All manufacturers collect data—whether with pen and paper or full-scale automated data collection. And regardless of the end goal, the operators, quality professionals, and supply chain managers are generally all told to collect more data. They’re told to check product quality, machine safety, personnel compliance, and inventory levels more often.
In theory, this increased data collection will ensure that you and your facility are ready for an audit, prevent release of inadequate products, avoid injuries, increase efficiency, maintain adequate supplies of materials, and send out all your orders on time.
While these checks are valid and important, the theory of more may be overkill. Really, it’s not about more data, but the right data and better analysis. The data you collect must have value, be meaningful, and be concise. Applying Statistical Process Control (SPC) methodologies to the right data will help you proactively control processes and prevent manufacturing quality problems before they occur.
2. Predict the Best Opportunities for Improvement
Applying this proactive, predictive strategy, a client of ours in the packaging industry identified its worst-performing plant and gave the plant manager an ultimatum to improve performance.
Until this point, the plant manager had used traditional quality practices to identify products that did not meet specifications while they were in production. But, to make a greater impact under the ultimatum, he upped his analytics game and used the same quality data that determined the facility’s shortcomings to obtain operational insight and identify areas for improvement.
Within six months, his facility became the top performing plant in the company and he quickly began reducing costs, recalls, and defects.
3. Apply Best Practices across the Enterprise and Watch the Savings Roll In
Instead of individually focusing on improvements in each plant, managers—both quality managers and plant-level—should look at aggregated data across operations to make enterprise-wide adjustments.
This approach can exponentially increase the financial impact of the improvements. For instance, one of our consumer goods customers eliminated overfill on 12 production lines in a single facility to save more than $250,000 in two years. When applied to multiple facilities, the sustainable annual cost savings exceeds $1 million.
Look Ahead to Find ROI
Manufacturers are interested not just in quality control but also in ensuring the whole plant is functioning at an optimal level—uptime, staff efficiency, timely measurements, and the best product possible.
With predictive analytics, it’s possible to not only improve manufacturing quality, increase equipment return on investment (ROI) and overall equipment effectiveness (OEE), and anticipate needs across the plant and enterprise but also enhance your brand’s reputation, outpace the competition, and ensure consumer safety.
Instead of focusing on why using predictive analytics is beneficial, maybe it’s time to start finding ways you can use it so your organization doesn’t get left in the dust.
Learn more about the success of our packaging customer—and browse other case studies—to see how your organization can also benefit from better analysis of your SPC data.