The Robotic BarrettArm to program itself for applications

Citation

(1999), "The Robotic BarrettArm to program itself for applications", Industrial Robot, Vol. 26 No. 2. https://doi.org/10.1108/ir.1999.04926bab.004

The Robotic BarrettArm to program itself for applications

The Robotic BarrettArm to program itself for applications

Keywords Robotics, Robots

Barrett Technology manufactures advanced robotic arms and graspers with superior end-user flexibility based on innovative technology.

A collaboration of research teams at the University of Massachusetts, Amherst, are using the BarrettArm and BarrettHand to simulate and understand infant learning. Robot programming, normally a long, tedious task, is a major road block for further growth in robotics. If a robot can learn instead of being programmed for each task, the consequences are profound.

The proposed research is based on two studies. The first study focuses on the learning behavior of human babies in their natural environment. The second study researches the nervous system's ability to learn and transmits signals for producing appropriate actions. The BarrettArm, Wrist, and Hand will be controlled using a muscle model and genetic learning algorithms to simulate infants' learning behavior.

The BarrettArm BA4-310 was chosen for its smooth movement and low inertia, allowing the arm to interact arbitrarily and safely with its environment, much like a human baby does. The BarrettArm (a.k.a. WAM arm) was developed at the MIT Artificial Intelligence Lab. The BarrettHand BH8-255 was chosen for its compact size, low weight, and versatility.

The research will be headed up by some of the finest researchers at University of Massachusetts: Professor Roderic A. Grupen, Professor Andrew H. Fagg, Professor Andrew G. Barto, Professor Paul R. Cohen, Professor Neil E. Berthier, Professor Rachel K. Clifton, and Professor Carol R. Beal.

For information contact: William Townsend, Barrett Technology, Inc. Tel: (617) 252 9000; Fax: (617) 252 9021. E-mail: wt@barrett.com