Deep Green, the pool-playing robot from the Robotics & Computer Vision lab at Queen's University has been making the rounds on other websites recently. Its robot arm uses a huge gantry to move anywhere over the table, and then a special cue stick tool to hit the ball.
It’s hugely impressive, but the size and complexity of the hardware puts it out of reach for most Makers [please -- someone prove me wrong!]. So, that’s why I was attracted to the second half of the video: The students developed an equally impressive augmented reality mode that works without the robot. By using a projector and a camera located above the table, it recognizes the position of balls. As the player moves their pool cue, the system projects the predicted trajectory of the balls right on the table. If you can hit the ball consistently, this could be a great help in visualizing & setting up complicated shots!
See also: Roomba Pool
via waxy
Read more | Permalink | Comments |
David over at Boing-Boing spotted this jaw-dropping video of an experimental robot arm doing things that scare the bejeesus out of me. David writes:
Researchers from the Ishikawa Komuro Laboratory at the University of Tokyo presented this incredible video of a high-speed robotic hand at the 2009 IEEE International Conference on Robotics and Automation. The laboratory’s Web site has many more videos related to this project, called Sensor Fusion. Sensor Fusion: High Speed Robots
For my money, the most impressive stuff is around 2:22, near the end.
Read more | Permalink | Comments |