After a decade in pure mathematics, Fabrizio has joined Orbotix in the startup world, engineering new computer vision algorithms to back cutting edge augmented reality games. Originally a firmware team member, Fabrizio still contributes heavily to the robotic control system, the on-robot sensor fusion algorithms and factory calibration systems. A compulsive problem solver with a programming problem, Fabrizio rarely sees the light of day. On those rare occasions when he’s not working on his hobby language project you will find him playing disc golf.
Could you present yourself and your society Orbotix in few words ?
At Orbotix our goal is to continue to blend the real and virtual worlds to expand the boundaries of what we expect from technology. With Sphero and our innovative software we’re exploring and defining mixed-reality gaming – the next evolution of entertainment. This completely new category challenges our preconceptions of how we interact with technology. Founded by passionate robotic engineers, Orbotix is working to bring the real world to your smartphone. Based in Boulder Colorado, the Orbotix team takes fun very seriously.
Without revealing the secrets, can you explain how does works the sphero?
Sphero is controlled via bluetooth. Inside the sealed polycarbonate shell is a two wheeled robot that moves by shifiting it’s weight like a hampster in a wheel. Our augmented reality technology tracks Sphero in video using a custom, high-performance tracker. Tracking data is combined with IMU and other sensor data from the ball to generate pose estimates for the camera and ball. All of this is accessible to developers through our SDK.
how do you imagine the interface between the robotics and the augmented reality?
Robotics and augmented reality work together marvelously. Just as the human brain seemlessly integrates felt limb position (proprioception) with binocular SLAM to estimate motion through space, Sphero AR fuses visual tracking data with robotic location information. Using some simple machine learning we are able to rapidly correlate robotic and visual data, combining them into a single, superior estimates for the position and orientation of both the ball and mobile device.
How do you personally imagine future use of this technology ?
Robotics is held back more by lack of intelligent software than it is by mechanics. Sphero AR represents an important steps toward integration of the two in a consumer product. In the future we should expect the marriage of robotics and computer vision to be ubiquitous. Embryonic examples include increasingly successful autonomous car projects like those participating in the DARPA Grand Challenge.