Our team is an alliance between MIT and Princeton University, where we blend expertise from manipulation with computer vision, respectively. The team is headed by Prof. Alberto Rodriguez, Assistant Professor at the Mechanical Engineering Department and PI of the MCube lab. The team members from MIT are MCubers and the team members from Princeton are advised by Prof. Thomas Funkhouser, PI of the Princeton Vision Group.
Team MIT-Princeton at APC 2016
Our team is composed of groups that each strive to achieve their respecitve goals while communicating and collaborating closely with other groups, below we describe the group objectives and introduce the members:
The objective of this group is to construct and maintain the software infrastructure for the project. This group strives to maintain a high quality of code content and contribution and provides software and support for other groups.
The objective of this group is to endow the robot with the ability to manipulate objects in the real world using its grippers and suction cups, reasoning about interactions and exploiting the sensory feedback to garner as much information as it can through the sense of touch.
- - - -
The objective of this group is to design and build the hardware necessary for the robotic system, most of the gripper, fingers, suction mechanism, and various attachements of the robot are in house designs and made by this team and undergo several rounds of design review.
Elliot Donlon - Melody Liu - Aparna Ramanathan
The objective of this group is the development of high level reasoning and decision making of the system, the robots ability to reason about which objects to pick, where to place, what sequence of actions to take, and which goal to consider at each given time.
Ferran Alet - Maria Bauza
The objective of this group is the visual perception of the robotic system, they use deep neural networks and various in-house developed algorithms to identify objects, localize them, and provide information regarding feasible regions of interaction.