In the past few years, we've seen a tremendous swing from metrics to analytics. However, a lot of people and companies are still confused about the difference. Or they think they know the difference but have a difficult time explaining it.
Metrics are a standard measurement that usually involves a count, e.g., number of days without an accident or number of pieces produced during a shift. For decades, process and production metrics have been the primary source of information to evaluate performance and show the shop floor how its systems are doing. For years, data historians and enterprise manufacturing intelligence have supported these applications.
Analytics don’t measure anything; instead, they answer specific business-related questions and, as the name suggests, involve analysis rather than just reporting of data. In the past five years, the manufacturing industry has seen a wide range of new analytics applications in asset performance management (APM) and other maintenance-related processes, typically focused on high-cost resources. Unsurprisingly, process industries with high-cost, high-risk, highly regulated operations, such as oil and gas and chemicals manufacturing, have led the way.
The difference is an important one: Metric tracking tells you what’s happening, while analytics allows you to figure out why things happen—and what to do about them. Analytics—when configured correctly—can add significantly more value than metrics, including shortening maintenance windows, improving throughput and quality, and generally having a positive impact on manufacturing outcomes.
Not all companies are moving at the same pace of adoption, however, and there are a variety of maturity levels within the global manufacturing space. To gauge maturity, LNS Research defined the stages of analytics adoption (shown in the table above).
The first step in preparing to move to broad-based analytics is building business use cases. Our research data shows that the use cases companies most often apply are the most straightforward and focused on high-cost items. The No. 1 use case is remote monitoring of presses, robots/automation systems and other expensive assets to optimize maintenance and reduce planned and unplanned downtime. Improving asset reliability falls in the same category.
By focusing on straightforward use cases, a company can easily forecast and track the ROI for analytics. This simplifies approval for pilot projects; when pilots hit or exceed targets, it’s a lot easier to get a green light from the C-suite to scale programs.
Businesses considering the move from metrics to analytics should prioritize these actions. First, analyze the company’s current use of metrics and build use cases—with specific goals—for each. Compare target use cases against those typical in manufacturing right now. If your industry uses high-cost assets, focus on improving utilization and reducing costs associated with those assets. All others should zero in on use cases to drive up revenue, improve the cost structure, or both. Even product quality and risk management are legitimate and powerful starting points for analytics.
Second, gather a cross-functional team to examine data silos and establish a data architecture to power analytics. Keep in mind that analytics will evolve as new technology emerges; a data-centric architecture will future-proof today’s efforts and investments.
LNS Research and MESA International recently examined the state of play across metrics and analytics in the industrial space. The report can help you evaluate where your company is compared to the rest of the market, and explore the architecture required for effective analytics.
>>Patrick Fetterman is vice president customer operations of LNS Research, managing company operations across new customer acquisition and customer success. He is also a research analyst providing collaborative coverage across the industrial value chain, including industrial analytics, manufacturing operations technologies, digital transformation and the Industrial Internet of Things (IIoT).