Its name is Cue, and if it challenges you to a game of H-O-R-S-E, you'd be wise to decline.
Built by a team of Toyota engineers in their spare time, the 6'3" humanoid has reportedly taken more than 200,000 basketball shots with near-perfect accuracy.
To put their robot to the test, the Toyota engineers enlisted a couple pro basketball players from a Japanese league to go head to head with the humanoid in a shooting contest.
As the video above illustrates, it was no contest at all.
The term "automatic" is often applied to human players who seem to effortlessly dial in on the basket every time.
It's a fitting way to describe mastery over a physical process with predictable variables. The physics involved in putting a basketball through a hoop aren't all that complex when factors such as distance, the weight of a basketball, and gravity can be precisely accounted for.
The video below illustrates how even a standard industrial robotic arm can accomplish the task.
Because the Toyota engineers designed their robot to shoot like a human, they had to deal with the added complication of backspin.
Human shooters rely on backspin, in part, to make any contact between the rim and the ball more forgiving. According to John Fantanella, author of The Physics of Basketball, backspin gives the ball added lift, allowing good shooters to shoot the ball at the slowest possible speed, resulting in a less violent rebound off the backboard or rim, which equates to a higher percentage of shots made.
The Toyota engineers haven't revealed the inner-workings of their robot, but at a minimum such a device would need some way to measure distance to the hoop, a task easily accomplished with an electronic rangefinder, as well as by more sophisticated sensors like stereo vision or LiDAR.
From there it's a matter of applying accurate force.
Modern electro-mechanical actuators are remarkably precise, meaning the Toyota engineers would only need to dial in an equation to yield the input forces necessary to send a spinning basketball along a defined trajectory.
The result is a robot that's automatic from the line.