A resource for safe and effective troubleshooting from the leaders in simulation-based training environment.

Gut feelings versus data-driven decision making

Safra Catz, CEO of Oracle Corporation, has a unique take on gut feelings. “Managers used to say, ‘I have a gut feeling.’ Do you know what a gut feeling is for a professional manager? It’s a pattern that they recognize. But if your system can recognize that pattern, if it’s not just a couple of managers who know that pattern, then the system’s gut feeling can tell you which way to go.”

In other words, managers who are making decisions based on gut feelings are in a sense data-driven—they’re just working with a really limited set of data from their personal experience. And if that’s the case, then it’s not such a big leap to making decisions based on large amounts of objectively verifiable data.

This is a fact that companies interested in surpassing their competition have already embraced. Top-performing companies use data over gut feelings to make decisions more often than companies that are not peak performers. McKinsey analysts found that companies in the top third of their industry with respect to using data-driven decision-making (DDD) were 5% more productive and 6% more profitable than their competitors who didn’t use DDD.

Other research has confirmed that adopting DDD leads to better performance. Larger organizations that adopt DDD practices early do best, but even smaller, single-establishment firms that are later adopters of DDD have a higher correlation with performance than companies that do not use DDD. This same study found that when you combine DDD with skilled workers and prior IT investment, there is a synergy that leads to higher performance than just DDD alone; and more DDD is associated with better performance.

So there’s ample incentive for companies to begin leveraging the data that they do have access to and assessing what technologies exist that can help them collect and analyze additional relevant data.

What does data-driven decision-making look like?

In practical terms, there are a number of ways companies can use big data to increase profits.

They may use it to create new business, as in the case of an airline that increased the accuracy of their arrival time data by using custom software to analyze available data including weather conditions, flight schedules, past arrival times, and radar station feeds rather than using pilot estimates, and eliminated the gaps between projected ETAs and actual arrivals.

Or, they may use it to drive sales, for instance, by generating instant promotions to online customers based on past purchasing and searching behavior.

Or, they may use manufacturing equipment sensor data to alert maintenance staff to developing problems in the equipment so it can be repaired during planned downtime before it breaks down and brings production to a costly, grinding halt. This strategy, known as predictive maintenance, is becoming increasingly popular among manufacturers.

Predictive maintenance

Predictive maintenance is maintenance that uses sensors built in to production machinery to detect problems in the early stages, before they become full-blown equipment failures that shut down the whole production line. If problems can be caught early enough, they can be fixed during scheduled maintenance or planned downtime, which is far cheaper.

Deloitte Insights reports that poor maintenance strategies can reduce a plant’s production by 5% to 20% and that unplanned downtime costs industrial manufacturers about $50 billion a year. Unplanned downtime results in higher maintenance, repair, and replacement costs. According to Emerson, a global business systems and solutions provider, fourth quartile companies (companies who rank in the lowest 25% among peer companies with respect to operations and capital performance) spend over 4 times more on maintenance than top quartile companies, who have made the commitment to leverage the data from sensors embedded into manufacturing equipment.

Sensors and PLCs are critical to predictive maintenance

Sensors are the cornerstone of predictive maintenance. Modern production facilities use motion, environmental, and vibration sensors. Motion sensors can be used to detect when some part of the machinery is tilting, to detect shocks or falls, or recognize linear or angular positions. Environmental sensors measure humidity, temperature, environmental gasses, performance of pumps and pressure lines.

Sensors that monitor vibration levels in the machinery are very common and used extensively in the food and beverage industry to monitor conveyors, pumps, fans, and motors. They can monitor rolling elements used throughout the paper and plastics and mining industries, but which are difficult for maintenance technicians to access and check on. They can monitor vibration in machine tool spindles and alert techs to the need for maintenance before damage occurs.

Sensors are also used to monitor fluid power systems and keep an eye on connectors, hoses and tubing, pumps, actuators and filters, and more.

In order for all of this sensor data to be used intelligently, the sensors must be able to communicate with central PLCs (programmable logic controllers). PLCs are industrial computer control systems that use a custom program to monitor multiple input devices (such as sensors) and then make decisions to control the output devices, such as sending alerts to maintenance staff.

Sensor and PLC repair

These smart sensors come with huge advantages for manufacturers—mainly the massive savings in production-line downtime. Like everything else, though, these high-tech sensors and associated devices such as PLCs do need to be repaired or replaced from time to time, requiring special training for maintenance staff.

Which brings us to… Simutech Multimedia’s newest training module, PLC Sensors. This exciting new module will train maintenance professionals about industrial PLC analog inputs, PLCs and associated devices (It’s extra exciting for us because it’s our first module to use an immersive, life-like, 3D environment. It’s set in a greenhouse—get a sneak peek here!) PLC Sensors is now  available for beta testing —be sure to sign up here if you’re interested in being part of the beta.

And that’s it for today, Troubleshooters. Tune in to TST next week, when we’ll discuss how to build your next training plan.

Follow Troubleshooting Thursdays

Tune in to Troubleshooting Thursdays for reliable tips, general troubleshooting process, and industry insights. Stay up to date with Simutech Multimedia:

Have a subject you would like Troubleshooting Thursdays to cover? Send Simutech Multimedia an email at [email protected].

Looking to give simulation learning a try? Get started with our award-winning first solution here: Get Demo.

Get started with your demo

Share This