Brain-Emulating Cognition and Control Architecture (BECCA) V1.0 beta

Software · Sun Sep 30 00:00:00 EDT 2007

OSTI ID:1231144

Rohrer, Brandon

BECCA is a learning and control method based on the function of the human brain. The goal behind its creation is to learn to control robots in unfamiliar environments in a way that is very robust, similar to the way that an infant learns to interact with her environment by trial and error. As of this release, this software contains two simulations of BECCA controlling robots: one is a one degree-of-freedom spinner robot and the other is a 7 degree-of-freedom serial link arm with a terminal gripper. In addition, the software contains code that identifies synonyms in a untagged corpus of ASCII words. This last is a demonstration of BECCA's ability to generate abstract concepts from concrete experience. The BECCA simulation is coded so as to make it extensible to new applications. It is modular, object-oriented code in which the portions of the code that are specific to one simulation are easily separable from those portions that are the constant between implementations. BECCA makes very few assumptions about the robot and environment it is learning, and so is applicable to a wide range of learning and control problems.

• Short Name / Acronym: BECCAV1.0 Beta; 002318WKSTN00
• Version: 00
• Programming Language(s): Medium: X; OS: Any; Compatibility: Workstation
• Research Organization: Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: DE-AC04-94AL85000
• OSTI ID: 1231144
• Country of Origin: United States