Artificial intelligence company Bonsai AI Inc. is showcasing how its technology can be applied to industry to help manufacturers become more profitable.

In collaboration with German conglomerate Siemens AG, Bonsai today revealed that it has created an AI model that can autocalibrate Computer Numerical Control machines 30 times faster than an expert human operator can do. The potential implications of this achievement could be massive, the companies said.

CNC machines, also called computer-controlled machines tools, are used in a wide variety of manufacturing operations. They revolutionized manufacturing in a wide range of industries following their introduction back in the 1940s, but like most technologies there’s always room for improvement.

One of the problems with CNC machines is that there’s a constant need to recalibrate them because of the wear and tear they suffer while being used on a daily basis. With such precise tools, it’s necessary to calibrate them to a high degree of accuracy to avoid errors that could lead to imperfections in the products being made. Traditionally, manufacturers have had to rely on specialist engineers to calibrate their CNC machines for them, which is both costly and time-consuming and entails a long period of downtime.

Bonsai, which leverages the open-source TensorFlow machine learning library that was first developed by Google Inc., realized this problem and decided its technology could be used to make the recalibration of such tools much more efficient. Working with Siemens, the companies used a technique called “machine teaching,” or deep reinforcement learning, to train CNC machines to perform complex tasks.

The project involved building a predictive AI model that would learn to calibrate CNC machines. Bonsai refers to this model as a Basic Recurrent Artificial Intelligence Network, or BRAIN, which is trained using deep reinforcement learning algorithms.

Within six months, Bonsai and Siemens had cracked the problem, creating a BRAIN that could calibrate a CNC machine 30 times faster and also to a much higher degree of precision than a human operator could do.

“Our successful project with Siemens represents a huge milestone in industrial AI, demonstrating the powerful results that can be achieved by combining machine teaching and machine learning,” said Mark Hammond, Bonsai’s co-founder and chief executive officer. “The beauty of this approach is that it balances the best of human and machine intelligence. Applied across the whole industrial manufacturing sector, the implications are staggering.”

Image: Bonsai AI