Ph.D. Prelim: Matthew Gardner

Event
Speaker: 
Matthew Gardner
Friday, October 11, 2019 - 2:00pm to 3:00pm
Location: 
1105 Pearson
Event Type: 

Title: Motion Estimation and Robotic Batting of Objects in Free Flight

Abstract: Robotic batting of an object to a target is a skillful task that requires accurate perception of the flying object, robust modeling of impact dynamics, and efficient planning of a robotic arm's motion. Leveraging of impact and measuring of motion are of great importance in manufacturing, sports, and space robots. To demonstrate the use of impact, we solve the batting problem in two dimensions based on impulse and energetic restitution with friction, flight mechanics incorporating gravity and aerodynamic forces, and trajectory re-planning for the bat-wielding robotic arm. Experiments with different objects show better batting performance than a human with no training.The component of estimating the pose and motion of an in-flight object is subsequently being extended to three dimensions. Typical robotics applications rely on large, complex vision systems that may be infeasible due to limitations of cost, space, power consumption, etc. We present a stereo vision system consisting of two high-speed cameras. A hypothesis-based algorithm is proposed to track the object's varying topology across the acquired image frames, and Kalman filters are employed, one for each currently active hypothesis, to compete for the estimation of linear and angular motions. Also under consideration is aerodynamics of the object as well as the two-view geometry enforced by stereo cameras. Preliminary results have been obtained from the flights of varying objects, and compared against calculations based on accelerometer data and image coordinates. Remaining work includes a comparison of modern Kalman filtering techniques, and extensive validation of the tracking algorithm and state estimation, especially in the presence of strong aerodynamic forces. The system, which currently assumes a convex polyhedral object, will also be generalized to shapes with curved edges.

Committee: Yan-Bin Jia, Sourabh Bhattacharya, David Fernandez-Baca, Guang Song, and Alberto Passalacqua