- Videos / Webinars
- Print Archive
- Events Calendar
R ecently more than 194 manufacturing IT managers and directors were asked to identify what would facilitate collaboration and productivity improvements across their organizations. More than 150 people completed a multiple-choice email survey, which contained both open-ended and multiple-choice responses. Of those who responded, 95 percent identified quality assurance management as a critical area that needed to improve technological collaboration.
Many of the respondents represented medical device, automotive assembly, and electronics manufacturing enterprises; others were involved in some aspect of machining. Although respondents perceived that the quality operation was siloed from other company functions, they noted it was simply part of the company’s culture rather than a deliberate isolation. This quality silo had evolved from a lack of technology, preventing collaboration across functions.
Traditional quality software is insufficient, according to respondents, because it is strictly measurement-based; it measures quality as a separate element of the part being manufactured and doesn’t relate it to the entire manufacturing process. Too many solutions treat quality as a separate function in the organization as well, so it doesn’t help with integrated problem-solving across teams or functions. Because of technological barriers, quality becomes a silo in the organization and doesn’t give staff a holistic view of the manufacturing process.
When evaluating metrics collected by dozens of quality management software solutions, too often the focus is on quality measurements only and not always on the use of quality equipment, such as test stations. There can be a 3:1 ratio of time spent in inspection vs. time in production. In other words, for every single unit of time spent machining something, three units of time are spent inspecting the product produced.
Although in many plants testing can be the bottleneck, monitoring and analyzing the use of quality assets is also critical. Yet, this process is rarely included in traditional quality management software products.
Another critical element is the ability to monitor whether test equipment is functioning properly. Without verification that machining inspection equipment is accurate (i.e., whether it’s measuring within the stated accuracy of the equipment), there is no way to know if test equipment has failed. Few quality management software solutions include this check and balance element.
What is required is predictive quality control, which facilitates examining quality data with knowledge of the process so that quality can be predicted using quality data.
Those process data must specifically focus on the key process parameters that drive part quality. Solutions must utilize deep historical data collection and pattern matching. The machine “learns” to identify patterns which become predictive. This ensures quality through process parameter control and adherence.
Manufacturers want to increase device use, reduce scrap rates, decrease unplanned downtimes, and improve profitability. By improving machine utilization and reducing energy consumption, many manufacturers realize savings between $30–$100K per machine, per year, as reported by firms such as Curtiss-Wright and Task Force Tips. Understanding the utilization of quality equipment is vital to recognize if more equipment is needed and to create best-practice optimized resource planning.
Many companies implement overall equipment effectiveness (OEE) solutions in an attempt to make best use of their equipment assets. However, more than three quarters (78%) of survey respondents report that little value was extracted from the data provided by their OEE solutions.
Athulan Vijayaraghavan, CTO of System Insights, explains that the OEE metric of availability is inadequate for predictive quality control. “OEE software solutions can tell you what happened, but not why,” he says. “They generate reports, but do not help make sense of the data. By applying real-time pattern-matching capabilities, our product—vimana—helps make sense of the data and illustrates how they can be applied in improving productivity and quality. By using dashboards, real-time metrics, alerts, and contextual reports, measurable value can be established within weeks of deployment. Operations, IT, and quality users need solutions that allow all to quickly understand the causes of production losses on the shop floor, including machine breakdowns, poor quality, material starvation, and incorrect machine usage.”
System Insights’ vimana software is one of the newer “big data” solutions. It can predict quality “spills” (i.e., false positives and negatives) based on process data, allowing quality assurance professionals to understand which parts are spending more time getting measured. The quality silo is eliminated because there is a unified application for process, quality monitoring, and analytics. The result is a collaborative decision-making process across horizontal teams.
Essentially, vimana connects directly to the software that controls and monitors CNC equipment and other process equipment, as well as inspection equipment. In the latter case, vimana pulls data directly from inspection software—Dimensional Measurement Interface Standard (DMIS), for instance. The ability to track both production and inspection data and to see the correlation between them is what gives vimana its predictive power. As long as vimana has access to direct machine data, it provides a way to correlate downstream process failures with upstream production or measurement steps, regardless of what equipment is being used.
In the case of monitoring inspection equipment, that scenario might play out as follows: A part that has already been passed by an inspection station might not work in a downstream process. By looking at all the data on that part and all parts that came through that inspection station, a determination might be made that the inspection equipment, a coordinate measuring machine (CMM) for instance, was out of calibration.
Conversely, a CMM could be monitored for trending data as an indicator of imminent machine failure in an upstream operation. In this case, the data analysis is used to predict the need for machine maintenance.
By monitoring key parameters on quality equipment, manufacturers can understand and predict test equipment failures. Beyond simply monitoring process data, quality data can be captured using MTConnect, a manufacturing industry standard that facilitates the organized retrieval of process information from numerically controlled machine tools. MTConnect is designed for the exchange of data between shop-floor equipment and software applications used for monitoring and data analysis, such as vimana. Optimizing the organization and planning manufacturing systems, the widely adopted MTConnect manufacturing data standard is compliant across a wide range of machine tools and manufacturing equipment. It is vital to collect both sensor data and monitor quality equipment as well.
Quality throughout the entire manufacturing organization is best served by providing machine data to identify the conditions and patterns that lead to quality deficiencies and, conversely, best practices. Precise pattern matching allows manufacturing enterprises to identify quality issues in real time. By monitoring all steps of the manufacturing process, both production and inspection steps, it’s possible to look at the manufacturing process holistically and connect what’s happening in one part of the process with what has happened, or will happen, in another.