Project Info

Learning Robot Motion Ability

Neil Dantam | ndantam@mines.edu

Robots in our homes will need to move through unstructured and even cluttered arrangements of objects. For example, consider a robot retrieving the ketchup from the back of your refrigerator. Can the robot maneuver its arm around the milk and peanut butter in front, or must it first remove those objects to reach its target in the back? The project will explore the ability or feasibility of robots to achieve motions by combining the search for valid paths with machine learning to classify whether the desired goal (e.g., reaching the ketchup) is feasible. The resulting algorithm will return either the valid path or a classification of the problem as infeasible.

For more information:

The Open Motion Planning Library (https://ompl.kavrakilab.org/).

The Task-Motion Kit (http://www.neil.dantam.name/papers/dantam2018tmkit.pdf).