8 years ago, The Pentagon’s Defense Advanced Research Projects Agency says robots slowly struggle (and sometimes fail) to perform a series of human tasks, such as opening a door, operating a power tool, or driving a golf cart. I organized a contest that was painful to do and watch. A video of them fumbling and tripping at the Darpa Robotics Challenge quickly went viral.
Today, the descendants of those unfortunate robots are much more capable and graceful. Several startups are developing humanoid robots that they claim could be employed in warehouses and factories within just a few years.
Jerry Pratt, senior research fellow at the Human and Machine Cognition Institute, a nonprofit research institute in Florida, led the team that came second in the 2015 Darpa Challenge. He is now the co-founder of Figure AI. A humanoid robot designed for warehouse tasks announced today a $70 million investment round.
Pratt said that if DARPA’s challenge ran today, the robot would be able to complete the challenge in about a quarter of the 50th it took his robot to complete the course, and with few accidents. I’m here. “From a technical standpoint, there are a lot of enabling technologies coming up these days,” he says.
Enabled by machine learning developments over the past decade, more advanced computer vision has made it far more possible for machines to navigate complex environments, climb stairs, and perform tasks like grabbing objects. made easy. The more power-dense batteries produced as a result of the development of electric vehicles also allow humanoid robots to pack enough energy to move their legs fast enough to dynamically balance. became. This means it can stabilize itself when slipping or slipping. Just like humans, sometimes we make the wrong step.
Pratt said his company’s robots are taking their first steps around a simulated warehouse in Sunnyvale, California. Figure Inc. CEO Brett Adcock believes that if there is enough demand to scale up production, it should be possible to manufacture humanoid robots for the same cost as manufacturing cars.
If Adcock is correct, the field of robotics is at a critical moment. You’re probably familiar with the dancing Atlas humanoid robot that has been garnering likes on YouTube over the last few years. They are made by legged locomotion pioneer Boston Dynamics, which built some of the humanoids used in the DARPA contest, and show that it is possible to build capable robots in human form. increase. But these robots are very expensive, with the original Atlas costing millions of dollars. It also lacked the software necessary to make it an autonomous and useful robot.
Figure isn’t the only company betting on the maturity of humanoid robots. Others include 1X, Apptronik and Tesla. Tesla CEO Elon Musk first visited his Darpa Robotics Challenge in 2015. The fact that he’s now keen to build a humanoid himself suggests that some of the technology needed to build such a machine is finally feasible.
Jonathan Hurst, a professor at Oregon State University and co-founder of Agility Robotics, was also at the DARPA Challenge to demonstrate his walking robot. Agility has been working on legged robots for some time, but Hurst says the company has taken a physics-first approach to locomotion rather than mimicking how human limbs work. The robot is humanoid but has ostrich-inspired legs.