David Niewinski of Dave’s Armoury won the ‘Jetson Project of the Month’ for building a robot arm capable of playing a perfect game of cornhole.
David Niewinski of Dave’s Armoury won the ‘Jetson Project of the Month’ for building a robot arm capable of playing a perfect game of cornhole. The robot runs on an NVIDIA Jetson AGX Xavier Developer Kit and can throw a perfect cornhole game.
For the uninitiated, Cornhole is a lawn game popular in the United States where players take turns using their aim and motor skills to throw bags of corn kernels at a raised platform which has a hole on the far side. Dave’s setup pairs the Jetson with a Kuka KR20 robot (fondly called ‘Susan’). A 1080p webcam serves as the eyes of Susan and a 2020 extrusion bar mimics the throwing arm of a player. The platform’s hole is colored red to make it easier for Susan to spot it from the background.
For the software, Dave used several OpenCV functions such as inRange to pick out the red hole from the scene, and findContours to establish the ring around the hole. Using the relative positions of the camera and the center of the hole, the angle and power for the throw are calculated on Jetson. Lastly, Jetson communicates these calculations to Susan through the network via the KUKA.ethernetKRL software package.
In the demo video, Dave mentions that he enjoyed working on Jetson and added,“This [Jetson AGX Xavier] is an awesome computer — think of it like if a video card had a baby with a Raspberry Pi. It has a lot of parallel compute on it, so you can do neural networks, deep learning, machine vision, but it doesn’t actually draw all that much power and with a little mount, you can strap it directly onto Susan.”
This project demonstrates how Jetson AGX Xavier could be used as an intelligent robot controller and can be paired with robot arms for industrial applications. Summer is here in North America and we’ll take some inspiration from Susan for our next cornhole game.
If you’re interested in learning more about the winning duo of Susan and Jetson, check out Dave’s code on GitHub.