Napoleon is reported as having said, “An army marches on its stomach.” Similarly, it can be said that a business marches on its data. In today’s rapidly moving marketplaces and swiftly changing technological landscape, data is the backbone that forms the basis of practically every business decision made. Without data, a company ultimately begins to sink.
It is critical for a company to have data available—but that data must be current, relevant, interpretable, and, most of all, reliable. Reliable information is the foundation of any production management system and the core of any continuous improvement project.
Accuracy versus Precision
Knowing whether or not your data is reliable depends on understanding two words: accuracy and precision. These words are often used interchangeably, but in fact, they have entirely different meanings. Accuracy refers to how close a measurement is to a pre-identified standard. For example, if you are weighing an object known to be exactly one pound, an accurate measure would be one pound.
Precision, meanwhile, has to do with measurement repeatability. Thus, if you weigh the one-pound object five times and obtain measures very close together, the measurements are said to be precise, even if they are not exactly one pound. In this case, the measures would be referred to as precise but inaccurate. Reliable reporting should produce data that is both precise and accurate.
Reliability of Manually Collected Data
Many production measurement systems use manual data collection. Technicians walk through the factory, manually recording data by counting product or recording information directly from equipment or production records. Analysts then process these records and create reports to present to management. However, there are inherent problems with this approach:
Human Error
Keyed data is prone to human error. In a study conducted in a medical setting, data keying accuracy was measured on a field-by-field basis with the following results: “Across all fields, the error rate was 2.8%, while individual field error ranges from 0.5% to 6.4%.” This error rate means that for every 100 bits of data entered, approximately three were incorrect. This level of mistakes can be expected in any manual system.
High Labor Intensive
Data collected manually is labor-intensive. After data is collected, it must then be converted into reports. Also, personnel on the production lines may be required to take time to make data entries of their own. Line personnel often consider these tasks to be a low priority, and, as a result, data accuracy is sacrificed in favor of a bit more production.
Slow Turnaround Time
Management and performance data value decreases over time. The sooner a manager knows a machine is running slow or is idling, the sooner he or she can fix the problem. Manual systems are characterized by chronically slow turnaround times, and supervisors must wait for results. Delays mean lost production.
Incorrect Measurement/Misinterpretation
The system must be designed so that only useful data is measured and reported. For example, for hand-cut steaks, should the team weigh the individual items, the usable scrap, the floor waste, the dimensions, all of these, none of these, or something else? How should checkweighers be calibrated? Every step in the process must be studied to identify the appropriate item or items to measure.
Are You Receiving Reliable Data from Your Production Line?
Knowing you are receiving reliable data is based on verifying measures, ensuring repeatability, making proper equipment calibrations, and training employees. Any measurement system can deliver reliable results if enough effort and money are spent managing the process. However, in today’s fast-moving business world, receiving reliable data a week or a month late is not acceptable.
Is Your Data Collection System Effective?
Data must also be obtained quickly and be presented in a format that is easy to understand and decision-centric. Usually, this means automating the process. For example, accuracy can be increased manyfold using barcode scanning, which delivers error rates of fewer than one error in one million scans. Quickly reporting results is also essential. Remote sensors attached to line equipment can send data wirelessly to computers, which then analyze the data and produce real-time results.
Interpreting the Data
Graphical or other data presentations should be designed to help managers make decisions about production line processes and efficiencies. Measures such as downtime or line slowdowns should show up immediately so corrective action can be taken. Smart Factory analytics instantly process measurement data and help managers make clear-sighted decisions.
Get Real-Time Results From Your Production Line
Reliability is an essential element in any production line measurement system. Getting the information you need when you need it is key to reaching optimum performance. Data that is collected automatically by equipment and checkweighers can be sent wirelessly to computers. Real-time results can then be produced, giving managers a profit edge in today’s highly competitive marketplace.
