Be Bold and Apply Data Analytics to Operational Data: Your KPIs will Thank You
2018-08-23 | 5 min read
There are two types of business assets: tangible and intangible. Your largest set of tangible assets is a major line item in your financials: plant, property and equipment (PPE), which includes test equipment. These days, the abundant data coming from that test equipment is among your largest intangible assets.
Such data delivers tremendous value because it can tell you what’s actually happening inside your operations—if you choose to listen to it. A few example use cases will illustrate what your data can tell you when data analytics (DA) converts it into actionable insights around key performance indicators (KPIs) such as yield, quality, throughput, utilization, and cost.
Exploring four specific examples
Many of our customers are in the early stages of applying data analytics for test and measurement (DA for T&M) to their operational data. As Bob Witte pointed out in his post, understanding your existing data helps improve overall knowledge of your business—and this enables you to define the key questions operational data can help you answer.
Among the Keysight customers who are actively climbing the maturity curve, many are applying DA for T&M in manufacturing. A majority of these projects align with one of four KPIs: warranty returns, mean time between failures (MTBF), test throughput, or quality and yield.
Reduce warranty returns
Let’s suppose one of your product lines is passing a finely tuned battery of tests in manufacturing; however, many are failing in the field and coming back as warranty returns. I suggest that you use DA for T&M to analyze data from every test in each step of your production process. Any outliers or problematic trends in the data will be visible, and you can correlate measurement results to every device by serial number. This will enable you to capture “walking wounded” devices—those that marginally pass or have been poorly reworked—before they can disappoint your customers.
As a regular practice, your staff may perform routine maintenance on crucial test equipment according to a fixed schedule. The upside: greater peace of mind. The downside: hours of instrument downtime, even though it is scheduled; and lost time for the people who are working on equipment that doesn’t actually need attention.
I would suggest a more efficient approach: Applying DA for T&M lets you shift from routine maintenance to preventive maintenance based on statistical, data-driven predictions of emerging issues or pending failures. This extends the mean time between failures and reduces downtime. It also leads to greater asset efficiency and utilization.
Accelerate test throughput
Looking across multiple lines that are manufacturing the same product, you may see significant variability in test times. Using DA for T&M, you can isolate those variations down to the exact test or measurement. This reveals actionable information about differences in test programs, and you can recommend changes that will optimize and accelerate specific tests or procedures. Taking this idea even further, one of our most advanced customers is improving throughput by applying basic machine-learning techniques to real-time data and making on-the-fly adjustments to test programs.
Improve quality and yield
Outsourced manufacturing adds complexity to many of your processes. DA for T&M opens the door to real-time process monitoring and control. For example, it can provide alerts based on variations in measurement data from specific components. This may reveal issues such as dual sourcing of components or accidental (or unauthorized) changes to test limits.
Taking the next step
Operating without DA for T&M is like driving in an unfamiliar city without a map app: you’ll eventually reach your destination, but you could have gotten there faster and with less frustration. Automated tools, dashboards and reports can guide you along the entire product lifecycle—and this applies to virtually every function, department and team within your operation.
Let’s discuss: What tools are you using? Which KPIs are you tracking? What sorts of improvements have you been able to achieve?