New types of robotic manufacturing equipment and industrial gear are likely to soon feature neural networks that can provide much more versatile grabbing, picking and sorting mechanisms than industrial operators are currently used to.
This morning, a Verge report covers robot arms on the ground near Berlin, Germany where a company called Covariant has deployed robots with multi-lens cameras and sensors, powered by machine learning algorithms, that can help the robotic arms equipped with suction cups to handle items of different shapes and sizes.
These new robotic entities can also, in an unprecedented way, pass on the knowledge that they acquire as they work.
“Thanks to a neural network it will one day share with its fellows in warehouses around the world, anything it learns, they’ll learn, too,” writes James Vincent. “Show this bot a product it’s never seen before, and it’ll not only work out how to grasp it, but then feed that information back to its peers.”
Vincent cites comments from installers about the possibility of quick deployment of these machines to manufacturing centers around the world.
He also contrasts Covariant’s processes with some of the biggest projects in robot arm grabbing, for example, the notorious Google facility fondly known as “the armpit” by insiders, where the “Don’t be evil” company is hard at work trying to get the same smarts out of robots.
Another takeaway from this story is that companies are being more choosy about how they approach these types of automation.
Citing hyped, glossy, short-form demos, Vincent suggests that many of these companies are seen to be ‘all hat and no cattle.’
”Our customers don’t trust short demo videos anymore,” says Pieter Abbeel, Covariant co-founder and the director of the Berkeley Robot Learning Lab, as quoted in Vincent’s explanation of new strategies in the manufacturing sector. “They know very well most of the difficulty is in consistency and reliability.”
To that end, expect decision-makers for industrial automation to really be doing their due diligence and researching how cutting-edge technologies work before making an acquisition or signing on for subscription services.
It’s an exciting time in manufacturing technology – many of these machines are valuable not just because they save money, but because they can replace various kinds of mind-numbing, back-breaking human labor that have plagued low-income workers for centuries. As long as we can figure out how to support relatively unskilled laborers and their families, these technologies could really improve our overall quality of life as societies.
