Robots Become Even More Productive With A.I, At CEATEC 2018

Most people think of industrial robots as systems that work at peak efficiency from day one, and eventually gets replaced for a new model, but at CEATEC 2018, many companies where showing how currently deployed system can benefit from the Deep Learning AI to perform even more.

During the manufacturing process, robots must move things around, assemble parts and run verifications. Japanese company FANUC was showing a robot that is capable of learning how to perform essential tasks much better.

In the demo, the robot was tasked with picking up odd-shaped parts with holes in them. Because these robots typically use an air suction system to hang onto the part, the holes make it extra difficult, and the machine needs to learn where to grab the parts from. That’s where the AI comes into play.

By trial and error, the system learns where the best spots are and what technique will work best depending on how each part is positioned at that moment. Progress is measured with a success rate percentage that increases dramatically when AI is switched ON or OFF.

We previously covered another arm-based robot that can tidy rooms, and this shows how progress in industrial robotics can start trickle into home applications. Because AI enables robots to perceive and learn, they can now perform tasks that are seemingly easy for humans, but previously nearly impossible for robots to perform reliably enough to become consumer products.

Future robots will utilize an array of different AI capabilities such as pathfinding to navigate, facial recognition to identify people (for customization or security), advanced motion to handle objects physically, and more. As they do, they will increasingly save time to their users, by doing things like fold laundry, tidy room, or perhaps even cooking (some day).

You May Also Like

Related Articles on Ubergizmo

Popular Right Now

Exit mobile version

Discover more from Ubergizmo

Subscribe now to keep reading and get access to the full archive.

Continue reading

Exit mobile version