Google LLC parent Alphabet Inc. today announced a new venture, Intrinsic, that will focus on making industrial robots easier and more affordable to use for companies such as manufacturers.

Intrinsic is a spinoff from Alphabet’s X research lab.  At any given time, the lab works on multiple emerging technology projects. When one of those projects begins to show significant promise, X spins it off into an independent unit that operates similarly to a startup, only under the Alphabet corporate umbrella.

Intrinsic was formed the same way. The spinoff started as an internal project at X about five and a half years ago. In that time frame, Intrinsic has developed a prototype version of its technology and tested it with a number of external partners. Now that it has been spun out of X, Intrinsic will work to complete development of the technology so it can be commercialized.

Intrinsic’s objective is to expand the use of industrial robots, such as robotic arms, by making them more accessible. “Intrinsic is working to unlock the creative and economic potential of industrial robotics for millions more businesses, entrepreneurs, and developers,” Intrinsic Chief Executive Officer Wendy Tan-White wrote in an announcement today. “We’re developing software tools designed to make industrial robots (which are used to make everything from solar panels to cars) easier to use, less costly and more flexible, so that more people can use them to make new products, businesses and services.”

In particular, the way Intrinsic hopes to make industrial robots more affordable and easier to use is by tackling a major cost associated with using them: programming. Training a robot to perform even a relatively simple task, like welding together two car parts, can take hundreds of hours of coding. The more tasks a company is looking to automate, the more software engineering work is required.

Intrinsic is tackling the challenge by developing artificial intelligence software that it hopes will significantly reduce the amount of time required to train robots.

The software uses multiple AI approaches under the hood including deep learning and reinforcement learning, according to Intrinsic. Deep learning is a technical term usually associated with the most complex and sophisticated AI algorithms. Reinforcement learning, in term, refers to a type of neural network that figures out how to perform tasks through a process of trial and error.

In one experiment, Intrinsic used its software to train a robot how to connect USB cables to a panel. The software enabled Intrinsic to train the robot in two hours instead of the hundreds of hours the process would have normally required.

In another project, the X spinoff partnered with a team at Swiss university ETH Zurich to build a sustainable architectural installation. Four ceiling-mounted industrial robots (pictured) powered by Intrinsic helped assemble parts of the installation. 

To help enhance its technology, Intrinsic is now looking to partner with companies in the auto, electronics and healthcare segment that are using industrial robots as part of their operations. The spinoff will also expand its team to help accelerate development efforts.

Alongside Google, Facebook Inc. is another tech giant investing in robotics software research. Earlier this month, a group of researchers from the social network, UC Berkeley and Carnegie Mellon University detailed a new artificial intelligence method that enables robots to navigate unfamiliar terrains more effectively than they could until now.  Earlier, Facebook released a development toolkit for creating AI algorithms that can operate in three-dimensional environments such as factories.

Similarly to Facebook, Google is no stranger to AI research. The company has an entire unit dedicated to developing new machine learning techniques, DeepMind, that achieved several major technical advances over the last few years. DeepMind’s know-how could assist the company as it works to establish a bigger presence in the robotics market through its newly launched Intrinsic venture and, possibly, any additional projects it may launch down the line.

Google made several attempts to target the robotics market in the past. In 2013, for example, the search giant acquired noted robot  maker Boston Dynamics, but sold the firm a few years later. 

Photo: Intrinsic