Using data to optimize industrial operations

Connected systems, with their ability to collect and crunch massive quantities of data, are transforming industry and business.

For one thing, they’re changing the role of flesh-and-blood human beings in industrial settings. For most of the long history of how we make things, humans have been forced to attend to minutiae—for example, to the question of what machine parts required replacement, and exactly when.

That’s coming to an end. Today’s tech is liberating humans to do the cerebral work machines will never be capable of. Machines, for their part, are taking over from people, and excelling at, the rote but crucial tasks—monitoring wear and tear on ball bearings and the like.          

Here are four ways that data is helping industry work smarter, faster, and with greater efficiency: 

Data-driven industrial methods keep developing. Condition-based monitoring (CBM) is a new maintenance trend that uses real-time data to improve on traditional predictive maintenance practices.

Predictive maintenance uses sensor-supplied data to discern with precision how a machine will function (or not function) in the future. In CBD, by contrast, real-time data enables the performance of repair and replacement work at the right moment. As soon as sensors detect that a generator is running hot, for example, the system prompts intervention.

Predictive maintenance forecasts the future, while CBD keeps an eye on what’s happening in the present. And as we all know, predictions are less reliable than observations.

A digital twin is what it sounds like: A computer-based model of a real-world object or system, in particular one of a size or complexity that defies the human ability to comprehend it as it is. Think of a nuclear reactor, a container ship, or even an entire city.
 

A digital twin puts that object or system in front of human managers, in forms that they can work with and study, from numerical datasets to 3D imagery. Those managers can, among other things, tinker with inputs, extrapolating from current conditions to test what might happen to the object or system given increased workloads or extreme temperatures.
 

Data, of course, is what makes the twin tick: The connected sensors that monitor the original item feed real-time information simultaneously to the twin.
 

One major strength of a digital twin is its ability to deliver information contextually—that is, according to users’ roles, use cases, and desired outcomes. As PTC business analyst David Immerman explains in a September 2019 article, a digital twin can deliver insights that go well beyond the repair and maintenance of industrial machines or systems. A service executive could use predictive maintenance to identify a product defect and take action to remediate it. An operations manager can use performance metrics to improve production processes. And an engineer can take the lessons learned from both to create better iterations of a product’s design.

Manufacturing and other industrial facilities are also getting safer thanks to data.

Sensor tech can trace how humans and machines (like forklifts) interact on crowded shop floors, generating data that can inform planning. In light of how employees move from their workstations to exits and rest rooms, what’s the best way to structure the floor? Which spaces in a facility remain underutilized, and which can be put to better use?

In the always-on environment of today’s automated manufacturing or logistics facility, answering such questions won’t just boost efficiency: it could also cut down on injuries and save lives.