Managing a production operation without productivity data is like driving a car at night without headlights. You can't see where you're going, you don't know where you've been, and you run the risk of suffering a dramatic accident. The sooner you're able to see and analyze your performance data, the better your decision-making will be.
Historically, either line personnel or individuals circulating throughout the factory collected operational and performance data. This information was given to analysts who calculated the required metrics. This approach to performance measurement was labor-intensive, introduced human error, and was limited to elementary measures. According to Mel O'Leary, CEO of custom blow molder Meredith Springfield Associates (MSA), "We used to rely on Excel spreadsheets and had 10 administrative employees, and the data were never accurate or up to date. Now we get real-time data, updated every two minutes." Faster response time for correcting problems and making line adjustments has resulted in double-digit growth for MSA.
In today's world of Industry 4.0, dark factories, AI, and IIoT, manual approaches do not provide the quality or accessibility needed to manage highly automated production operations. Automation, internet, AI, remote sensors, and robotics are driving a revolution in all types of production. Products are moving through manufacturing operations at higher speeds, with fewer human interfaces and improved quality. To stay abreast of these rapidly developing changes, factory managers today must look for better and faster ways to gather and interpret production data. Fortunately, recent developments in data collection techniques and software analysis allow more and better data to be delivered to managers faster, with improved analytics and more useful presentations.
Huge quantities of many kinds of data can be collected today, causing processing difficulties or data overload. Too much data leads to interpretation paralysis and a failure to identify relevant information quickly enough to act. As a result, factory managers must first look at what types of data are needed. As a manager gains knowledge, additional data might be collected and stored to be used for ad hoc analysis or future review.
It's essential to gather the right information at the right time. For best practices, data should be collected in real time, meaning at the time things are occurring. This typically means employing checkweighers, online sensors, or components attached to equipment and communicating wirelessly to the monitoring system. This approach yields instant data analysis and immediately highlights existing or developing problems. For example, machine slowdowns that might not be otherwise noticed can be immediately identified by analysis software and quickly investigated.
Initial performance monitoring should start with twelve basic metrics:
- Manufacturing cycle time
- Time to make changeovers
- Capacity utilization
- Overall Equipment Effectiveness (OEE)
- Schedule or production attainment
- Percentage planned vs. emergency maintenance work orders.
- Customer rejects/returns
- Supplier quality incoming
- Customer fill rate, on-time delivery, perfect order percentage
In addition to operating metrics, condition-based monitoring (CBC) can be used to gather other critical equipment functional data. Operating information such as vibration tracking, heat patterns, and acoustic analysis are just a few of the many machine data that can be captured. This data is used for predicting equipment breakdowns or maintenance requirements. Immediate feedback via wireless systems provides a real-time look at current and past performance. When deterioration in equipment function occurs, corrective action can be taken immediately.
Operating performance data captured in real time is essential for successfully managing today's increasingly automated production lines. To satisfy demanding customers, information such as line speeds, downtime, throughput, yields, and OEE is key to keeping a plant profitable.
In today's world of increasing automation, greater competitive pressures in the marketplace, and smarter factories, managers cannot afford to let the rest of the world evolve while they hang on to outdated, slow, manual methods of performance measurement. Leapfrogging to the future requires adapting connected automatic sensors to equipment and wirelessly collecting and analyzing key performance metrics.
Today’s manufacturing manager should understand the importance of real-time productivity data and the necessity of rapid action based on incoming statistics. Quickly analyzing the performance of each production unit and, if needed, raising that performance to acceptable levels is key to keeping a company competitive.