As a child, I often dreamed of having a magic wand—a tool that could transform my bicycle into a motorcycle, turn every vegetable into a cherry pie, and perhaps most ambitiously, give me my very own computer. I grew up in an era when computers were housed in data centers, requiring raised floors to manage the complex web of cables. Today, when I read about some of the Smart Manufacturing solutions out there, I get the feeling that these companies must’ve had a magic wand. They completely overlook one of the most challenging and essential tasks needed in building a Smart Manufacturing system: Data Conditioning. It’s unfortunate that all data is not created equal.
It comes in a variety of formats, sizes and nomenclature. Consider tank levels for example–whether its glue tanks, water tanks, medicine tanks, etc. manufacturing plants are filled with them. The tank levels can be measured in feet, inches, millimeters or even a percentage. Sometimes, it’s analog and the output of a sensor that generates a 0-5V, 0-10V or 0-24V signal.
If it’s digital, it might be expressed as a floating point, an unsigned integer, binary coded decimal value or even a string of ASCII digits. The challenges of data conditioning are inevitable. Equipment comes from different manufacturers with different standards and these standards are different all around the world. Since these kinds of problems are unavoidable, there’s no doubt you will have to deal with them sometime. You can deal with them at the equipment level.
You might do data conversion at the controller level. Or you might format your data in an aggregator. You’ll want to do it as close as you can to the original data source. The worst place to deal with data conditioning is in the Cloud where the source equipment might be obscured. Regardless of the system in place, data conditioning is unavoidable.
It’s an unfortunate aspect of Smart Manufacturing that a lot of vendors gloss over. While the allure of seamless Smart Manufacturing solutions can make it seem as if a magic wand is at work, the reality is that data conditioning is a fundamental and often overlooked challenge. Without addressing the diverse and complex nature of data—whether it comes from different equipment standards or varying measurement formats—the full potential of Smart Manufacturing cannot be realized. By tackling data conditioning early and at the source, companies can ensure more accurate, reliable, and effective systems. It’s a critical step that no vendor or manufacturer should ignore.
