The rise of artificial intelligence or A.I., has people questioning the future of everything from education and jobs to cybersecurity and robots.

Now, researchers at the University of Kentucky are figuring out how I can be used in manufacturing.

The group is focused on machine learning instead of telling a machine specifically what to do.

Hey, AI helps machines figure it out on their own.

We try to apply machine learning to manufacturing fields, but we try to make it in a much more explainable and deployable manner than just simply having all these complex equations, these really complicated models.

We want to actually see these things out in the world deployed.

And the only way to really do that is to make the models trustworthy or make them understandable to a normal person.

This is a pilot project related to a robotic assembly.

As you can see here, we try to apply robot to assemble multiple projects together.

The goal would be try to reduce and minimize that dimension with the assemble.

And we have our computer vision system.

So through the cameras you find the cameras according to cameras.

We can take a picture of what's happening on the table continuously.

And if you like to watch the image processing and the machine learning algorithms, we can identify how many objects are on the table, where all the objects and the shapes and dimensions, how objects and that information we all pass down to the robot.

So the robotic price motion accordingly and pick the objects one by one and assemble them together.

If the robot cannot accurately understand that information or only utilize that information, it will also lead to a failure that the robot can not to pick up the object or pick up that, but not successfully.

And we will design serious failure detection mechanisms and try to help the robot reliably and robotically to finish the work.

So the Google, we're integrating AI and machine learning solutions in manufacturing ways to help humans make better decisions or to fully automate manufacturing process or systems.

The whole point of manufacturing is high output, high quality redundancy, and there's still a ton of things that are manually done, either quality control or in or performance or assembly operation, so and so forth.

Well, the more manual labor you need that, the less and less consistency you get.

That's just the factor of humanity.

And so by incorporating machine learning, we get this more concurrent workflow and more stability.

So that's kind of what we're looking at as these novel ways of producing email to traditionally where humans are involved.

So if we can kind of cut out some of those those human aspects, we can, you know, get our process planning better and overall more stable.

Dr. Wong and his students are collaborating with companies like General Motors and General Electric Electric to access manufacturing plant data to bridge the gap between their laboratory simulat