magic-wand Getty Images

How the Magic of Predictive Technology Conjures Real Results

Downtime is a real problem in manufacturing, but the power of predictive technologies can make these issues disappear.

The ability to predict the future sounds like a comic book hero's super power or a smoke-and-mirrors act from a carnival sideshow. Yet, plant managers are expected to possess this skill every day when juggling resources, demands and operational pressures. Insight into the future lifecycle of each piece of machinery and equipment in the plant would certainly make planning easier. For those managers lacking mystical powers, predictive technology is the next best option.

Unforeseen downtime due to machinery-related issues can be a cause of colossal-sized headaches for the organization's COO, plant manager or shift foreman. Whether managing one assembly line or a global network of suppliers and fabricators, managers face unrelenting pressures to keep machines operating at peak performance with minimal interruptions.

Now you see it; now you don't

Even the smallest of broken fan belts, punctured coolant lines, or clogged filters can trigger a cascade of bottlenecks on dozens of interdependent processes. In a flash, smoothly humming processes can transform into chaotic, knee-jerk responses as operators redirect lines, shut down stations, answer warning alarms, and follow scripted safety protocols. When the machinery involves hazardous or dangerous materials, like molten metals or chemical coatings, the rush to respond is even greater as the impact can be potentially devastating. Such breaks from routine have become all-too-common in some facilities, especially when infrastructure is stretched thin and capital for replacing tired machinery is limited.  

Pulling costs out of thin air

Analyst firm Aberdeen Research recently estimated that 82% of industrial companies have experienced unplanned downtime over the past three years. When calculating the associated labor costs, lost productivity, overhead, and lost business, the financial impact becomes staggering. Aberdeen projected the true downtime cost (TDC) could be as much as $260,000 an hour.

The hidden costs can be difficult to itemize and, therefore, hard to quantify and calculate Return on Investment (ROI) for possible solutions. A recent survey conducted by Vanson Bourne asked plant managers how unplanned downtime impacted them. Responses indicate:

  • 45% could not delivery goods or services on-time
  • 37% lost production time on a critical asset
  • 25% said the downtime posed a security threat
  • 21% blamed downtime for lost business.

Understanding the financial ramifications—as well as the damage to customer relationships—is the first step in setting a strategy for prevention. With modern technology, plant managers no longer need to accept unplanned stoppages as an inevitable part of their world.

While every piece of equipment with wires, gears, belts, bearings, and or other moving parts will see fluctuations in performance and require preventive maintenance or repairs—those lifecycle "episodes" tend to follow a predictable pattern and provide a fair amount of warning signals. Just as the sniffles and a scratchy throat are early symptoms of a cold, industrial machinery provides early signs of malfunction.  For each asset the tell-tale symptoms will be different, from a spike in energy use to excessive vibration, overheating or gumming of adhesives.

The challenge is identifying those warnings and developing ways to monitor and detect slight nuances and variances early enough so that interaction can be planned during non-peak periods. The earlier the trigger incident is in the process, the more leeway there will be for appropriate responses, such as reserving placement parts, scheduling a technician, reassigning workflows, and determining if promised delivery dates need to be adjusted.

Technology to the rescue

Modern technology helps simplify and automate such processes, drastically reducing the occurrences of unplanned downtime—or the severity of impact. Here are five modern technologies that can help keep assets running as needed.  

Internet of Things: Smart sensors embedded in machinery capture and send performance-related data to the cloud to be aggregated and analyzed, looking for anomalies, such as temperature points outside of acceptable boundaries.

Artificial Intelligence: Business Intelligence (BI) solutions with AI built in determine the seriousness of the flagged data points and if actions are needed -- such as if a technician should be dispatched during non-peak hours or if a production line should be immediately shut down due to imminent danger to workers. AI can streamline complex decision-trees, apply data-based science, and proceed with defined actions, based on parameters learned from user input.

Machine Learning: As more and more data is collected and humans provide feedback that overrides or verifies the system's advice, Machine Learning capabilities help the system refine interpretations of data.

Predictive analytics: Not only will a BI solution analyze past and current data points, it can also apply predictive analytics to extrapolate what is the next likely data point or occurrence. These predictions are based on scientific algorithms. Machine Learning helps the predictions improve accuracy. Predictive analytics can project the lifecycle of equipment, suggesting when preventive maintenance should occur and when replacements of components—or the entire machine—should be planned. This window into the future makes it possible to plan for preventive maintenance and budget for replacements when needed.      

Preventive maintenance: A modern Enterprise Asset Management (EAM) solution includes features and functionality to help the service operation plan and track preventive care of assets so that the organization can be proactive in replacing items like belts and filters and replenishing fluids like coolant and lubricants. Preventive maintenance can also be worked into the schedule to help balance the use of technicians' time.

These five technologies can be instrumental in changing the occurrence of unplanned downtime. With the ability to predict the upcoming issues and identify issues early, the maintenance team can be proactive and take action before the issue escalates and become catastrophic or costly. Unscheduled downtime can have major impacts—from wasted resources to disappointed customers. Technology can make those headaches virtually disappear. 

For more information, visit

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.