|Covariant’s AI-powered robot can pick up and pack some 10,000 different items with accuracy greater than 99% / Credits: macor via 123RF|
Warehouses across the world are investing in robots powered by artificial intelligence to help them to automate tasks and make things easier for the employees. Recently, Californian AI and robotics startup Covariant operated its AI-powered robot at Berlin, Germany. One thing that makes it unique compared to other machines is that its robots can learn to pick things up. They can work 24 hours a day and pick items without fuss.
According to The Verge, an American technology news and media network that publishes news items, long-form feature stories, guidebooks, product reviews, and podcasts, Covariant’s AI-powered robot can grab and pack items with the help of a six-lens camera array and machine learning algorithms. They can also share what they learn with their fellows in warehouses around the world through a neural network. The startup also claimed that its machines can pick up and pack some 10,000 different items with accuracy greater than 99%.
“We tested this robot for three or four months, and it can handle nearly everything we throw at it. We’re really going to push these onto the market. We want a very high number of these machines out there,” Peter Puchwein, vice president of innovation at Knapp, the logistics company that installed the robot, said.
Puchwein added that their engineers traveled around the world to find the best picking robots. “Non-AI robots can pick around 10% of the products used by our customers, but the AI robot can pick around 95% to 99%. It’s a huge difference,” he said.
Covariant trained its robots using a variety of AI methods. The training is done in simulations, where the machines can take their time, often racking up thousands of hours of work. The company calls this result “the Covariant Brain” — a nickname for the neural network shared by the company’s robots. But, Covariant isn’t the only company using these technologies. For instance, startups such as Kindred and RightHand Robotics use similar fusions of machine learning and robotics.