Disney Research Pittsburgh has just released the video below which demonstrates one of its latest projects: an audio animatronic robot that can interact with people by playing catch with them. The system uses an off-the-shelf Microsoft Kinect (according to the video’s narration) along with an external camera system (ASUS Xtion PRO LIVE) to locate balls and a Kalman ?lter to predict ball destination and timing. So not only is the robot able to track a human’s position and size by the location of their head, but it can attempt to move its hand to catch the ball. If the robot misses the catch, it’s fully aware and even responds with one of several different humorous animations to elicit a response from the person interacting with it.
Disney Research has also been able to use its system to successfully juggle up to three balls at a time when a professional juggler is used as the participant.
According to the video and its description, Disney Research is hopeful that this product leads to a fully interactive experience between guests and audio animatronics in environments such as theme parks, while managing to keep the guest a safe distance from the robots.
In this day and age in which 3D scans of human faces are turned into exciting keepsakes such as the Disney/LucasFilm Star Wars Weekends experience ‘Carbon Freeze Me,’ in which guests could receive a replica of themselves frozen in carbonite a la Han Solo, and the upcoming ‘I Am A Princess,’ which builds on a previous test in which guests could have a princess doll in their likeness made, technology is becoming a key player in what has been even the most traditional of trades.
Therefore it shouldn’t come as a surprise that one of the pioneering technologies employed by The Walt Disney Company is being updated in a fascinating new way that will attempt to make audio animatronic figures rival the most advanced 3D, high definition screens. The ominous-sounding ‘Physical Face Cloning’ patent application (US 2012/0185218) seeks to improve upon the decades-old theme park experience by using some complicated algorithms to produce the most life-like audio animatronic figures to date.
Based on the listed location of the majority of the inventing team, the project appears to come out of Disney Research in Zurich, Switzerland. Disney Research recently came into the limelight with its technology dubbed touche, allowing users to control devices by gestures.
The problem with today’s audio animatronic figures, according to the patent, is that they require enlisting a team of animators, sculptors and other experts to create a face and skeletal system able to produce a realistic set of human expressions, as relayed through a layer of artificial skin. Taking the guess work out of the process, the new system could simply use motion capture technology to record the human subject’s face making various expressions and, via some very non-simple mathematical formulas, generate the perfect layer of silicone rubber skin (or whichever material is desired) of varying thickness, along with directions for attaching said skin to the skeleton, so that when the skin is stretched and manipulated on the figure to form the desired expressions, it provides the most realistic visuals possible.
UPDATE #1: 7/25/12 – Disney Research will discuss the new technology at SIGGRAPH 2012 on August 9. More information here.
For five days at Epcot in 1999 and for an equally short time in Castaway Cay, guests from around the world had the chance to meet with one of the largest accomplishments in audio animatronic technology: the DRU-1 (Dolphin Robotics Unit). Created by Edge Innovations in partnership with Walt Disney Imagineering, DRU-1 wowed the crowds but was ultimately decided to not have the potential of being an every-day attraction in the theme parks.
Here’s a video of DRU-1 in action courtesy of his WDI show producer:
And for more a lot more great information on the Walt Disney World project told from the perspective of someone who was directly involved, check out DisneyShawn’s blog here.