This robot can mimic your hand gestures and whoop you at rock-paper-scissors

From smart speakers like the Google Home and Amazon Echo to the Roomba autonomous vacuum cleaner, there are a growing number of smart A.I.s and robots we can call on to perform different tasks in our homes. A new collaboration between innovation studio Deeplocal and Google’s Internet of Things (IoT) framework Android Things wants to add another robotic helping hand to the mix — and we mean that quite literally.

Called HandBot, it’s a D.I.Y. robotic hand which can recognize your hand gestures and mimic them back to you, or compete against you in a classic game of rock-paper-scissors. To do this, it uses some smart machine learning-based image recognition, courtesy of an built-in camera that feeds it images of your movement.

“We teamed up with Deeplocal to build a series of demos to help inspire and show what developers can build by harnessing the power and potential of Android through the ease of the Android Things platform,” Melissa Daniels, a program manager at Android Things, told Digital Trends. “These demos also demonstrate the on-device processing power that makes Android Things unique.”

The HandBot robot is built using the Android Things developer kit, servos, and some custom cut acrylic. The palm of the hand contains five servos which move the fingers, and one servo for the wrist. The base, meanwhile, contains two more servos for forearm movement and other electronics such as an LED ring, PicoBoard, and the camera.

If you’re interested in creating a HandBot of your own, you can access all the open-sourced code on github and hackster.io, along with instructions for building it. The parts should set you back around $490, plus the price of an Android Things starter kit. The estimated build time is around seven hours, meaning that — provided you’ve got all the pieces — you should have no trouble building this over a weekend.

Sure, it’s probably not going to be the most useful gadget you’ve got on your shelf, but it’s a pretty neat way to dust off your engineering skills, while learning a bit of Tensorflow machine learning in the process. You can’t say fairer than that!

Editors' Recommendations