What is Machine Learning in Manufacturing?

Paul Siebert, Senior IT Analyst for our Manufacturing Operations practice, talks Machine Learning and its application in both our daily lives and the industrial setting. If you want to learn more, contact us at info@indx.com.

Full Video Transcript

Hi. My name is Paul Siebert, and I am a Senior IT Analyst. What is Machine Learning? Machine Learning is the idea of teaching machines how to think. A related concept being Deep Learning, which is specifically teaching machines how to think more like humans.

With Machine Learning, we are trying to apply mathematical concepts to allow machines to make decisions about future outcomes. We take some past data, we feed it into a statistical model, and we make a decision about what might happen in the future.

I guess the easiest example that I can give you of Machine Learning and Deep Learning is in your iPhone. Your iPhone learns the things that you do. It learns important locations, for example. If you open up your map, it'll tell you…You type in a location, and it'll say, all right, this is an important location because you've gone there a few times. And that's taking past data that you've put into your iPhone and modeling future events. I know that this person is likely to go to this location because they've been there so often in the past.

There are lots of different examples of Machine Learning and Deep Learning in your day-to-day life that you probably don't even recognize. It's very transparent, but the basic idea boils down to that.

How can we apply Machine Learning in manufacturing? Well, the truth is that we are already starting to apply machine learning in manufacturing. There are lots of different places that we can apply it. We can apply it for things like predictive maintenance. We can take in data directly from the shop floor as it is happening and make a decision on whether or not a machine needs maintenance. And this is all based on Machine Learning and Deep Learning.

We can use Machine Learning to do things like predict what machine is the most likely to produce a defect or how many defects is a single machine likely to produce. Those kind of things obviously provide great information for people to make decisions in the future.

The truth is that we're starting to see the beginnings of applications of Machine Learning and Deep Learning right now. It's being put into factories as we speak. It's going to change the industry. It's going to revolutionize the way that we produce products. It's only a matter of time.