CSAIL Research Abstracts - 2005 link to http://publications.csail.mit.edu/abstracts/abstracts05/index.html link to http://www.csail.mit.edu
bullet Introduction bullet Architecture, Systems
& Networks
bullet Language, Learning,
Vision & Graphics
bullet Physical, Biological
& Social Systems
bullet Theory bullet

horizontal line

A Design for a Self-Assembling Robot Arm

Myunghee Kim & Jeff Weber


We are trying to build a modular arm for a self-assembling robot. The design is focused on achieving modularity and simplicity to get the robot to manipulate objects in diverse environments. In addition, the robot also needs a force control to be able to manipulate these objects.


In the future, robots would build other robots especially in a harsh environment like the moon or Mars [1]. Modularization and simplification mean that those robots are easy to assemble and easy to produce. In addition, since we do not know the conditions of such an environment, the robots must also manipulate well under any conditions.


This robotic manipulator needs a device called the Series Elastic Actuator (SEA), which adds a capacity for active and passive compliance [2]. Therefore, we are building a force sensing load cell with an innate spring to ensure intrinsic compliance. In addition, based on the load cell, we are attempting to make a modular robot arm unit which has a load cell, gear, motor, position sensor and electronics that is as light as possible because a light but high torque arm makes manipulation easier and protects the arm from injury. Furthermore, a simple assembling method is also being studied to make assembly easier. We are also designing an electronic network to support our goals of simplicity and modularity.


After building compliance modular robot arm, we will try to get the robot to manipulate objects based on vision. At first, the robot will build Lego assemblies; then the robot will be able to create another robot which looks like itself while looking in a mirror.


[1] Robert A. Freitas Jr. and Ralph C. Merkle Kinematic Self Relplicating Machines. Landes bioscience. 2004.

[2] Aaron Edsinger-Gonzales and Jeff Weber. Domo: A Force Sensing Humanoid Robot for Manipulation Research. In Proceedings of the 2004 IEEE International Conference on Humanoid Robots. Santa Monica, Los Angeles, CA,2004.

horizontal line

MIT logo Computer Science and Artificial Intelligence Laboratory (CSAIL)
The Stata Center, Building 32 - 32 Vassar Street - Cambridge, MA 02139 - USA
tel:+1-617-253-0073 - publications@csail.mit.edu
(Note: On July 1, 2003, the AI Lab and LCS merged to form CSAIL.)