The Robotics Program at OSU covers a broad array of areas, from legged locomotion to human factors related to robots in the household. This particular project is focused on grasping - how to improve the ability of robots to pick up and manipulate objects. Applications include building better manipulators, designing tools that help robots manipulate everyday objects, and "learning" from humans how best to grasp objects. The goal of this project is to capture human insight into the best way to grasp and manipulate a variety of common objects, and to describe that insight in a form suitable for applying machine-learning algorithms. Example questions include: What is the best viewpoint from which to evaluate a grasp? Which hand positions are best for which classes of objects? What range of positions is a grasp good for? Depending on the intern’s skill set, s/he will also have the opportunity to apply machine learning to the data set or program a simple interface to elicit more precise feedback (such as where the hand should be positioned.) For more on our work, please visit http://mime.oregonstate.edu/research/rhcs/ and http://web.engr.oregonstate.edu/~balasubr/pub/Balasubramanian-et-al-ICRA-2010.pdf and https://ras.papercept.net/conferences/scripts/abstract.pl?ConfID=119&Number=183.
The mentor has a strong preference for applicants with one (or more) of the following: Basic programming skills, basic tool design (e.g. solidworks), basic electronics (e.g. working with a microcontroller). Female, minority, low-income and first generation college-bound students are strongly encouraged to apply for this position.