Alphabet has started up a replacement robotics company, Intrinsic, with Google’s parent firm looking to form smarter industrial ‘bots. The announcement comes shortly after Alphabet sold Boston Dynamics – known for its attention-grabbing human and animal-inspired robots – to Hyundai. Still, it seems Intrinsic will specialize in a different area of the sector.
While Boston Dynamics tackled both software and hardware for its robots, like models just like the dog-like Spot, Intrinsic will leave the physical robotics to others. Instead, it’ll specialize in the software they run, leveraging Alphabet’s expertise in machine learning, AI, and more to so on make existing industrial robots more flexible.
Though it will not seem quite as glamorous as making bipedal robots that will dance, it’s arguably an entire lot more useful for fields like manufacturing. “We’re developing software tools designed to form industrial robots (which are wont to make everything from solar panels to cars) easier to use, less expensive and more flexible,” Wendy Tan-White, new Intrinsic CEO, says, “so that more people can use them to form new products, businesses, and services.”
The challenge, Tan-White explains, is that while industrial robots – just like the vast robotic arms that help piece together cars on production lines and their far smaller counterparts which will assemble electronics – are capable, the method of programming them is often highly complex. That creates both the initial task of getting them to perform employment time-consuming and challenging and presents hurdles when it involves reprogramming them for brand spanking new jobs.
“Specialist programmers can spend many hours hard coding robots to perform specific jobs, like welding two pieces of metal, or gluing together an electronics case,” the new CEO says. “And many dexterous and delicate tasks, like inserting plugs or moving cords, remain unfeasible for robots because they lack the sensors or software needed to know their physical surroundings.”
Intrinsic‘s goal is twofold, then. On the one hand, it’ll be developing software that will use automated perception, deep learning, reinforcement learning, motion planning, simulation, and force control to teach robots how to perform tasks intuitively. On the other hand, it’ll check out how motion planning software can do training and deploy those enhancements more straightforward to actual businesses at an equivalent time.
“In one instance, we trained a robot in two hours to finish a USB connection task that might take many hours to program,” Tan-White says. “In other tests, we orchestrated multiple robot arms to assemble an architectural installation and an easy piece of furniture. None of this is often realistic or affordable to automate today – and there are many other examples like this in businesses around the world.”
While Intrinsic could also be breaking cover today, it’s not exactly a replacement startup. The team has been working within Alphabet for five and a half years, apparently, as a part of the X “moonshot factory” for brand spanking new projects. Now, alongside the general public reveal, it’s trying to find new partners – initially among the automotive, electronics, and health care industries – who want to explore how their existing industrial robots might be improved.