Research

My current research is focused on adaptive systems, specifically the development of cognitive processes and intelligent behaviors from the viewpoint of complex non-linear dynamical systems.  Non-linear and complex systems viewpoints are relatively new to cognitive science.  This view of cognition provides a powerful alternative to standard physical symbol system (symbolic) perspectives, as well as significantly extending connectionist/parallel distibuted processing views.  My interests and research touch on many related “movements”, such as situated and embodied cognition, dynamical systems, and epigeneitc or developmental perspectives.

A related interest is in computational science, also known as scientific computing.  I have always used large clusters of computers to build and evolve models of neural control systems.  My background is in computer science, and thus I have picked up some expertise on the practical side of using high performance computing clusters for such applications.

My focus is on building models of intelligent behavior from the complex systems perspective.  As a platform or application I use autonomous robots, simulation and enterprise level systems.  These models form the basis of planning and decision making abilities and our goal is to provide a means of general-purpose (non-task specific) autonomous intelligence for various artificial systems.  More detailed summaries of various projects and aspects of my work can be found by clicking on the summaries below.

Corvus

Building a curious cyber-infrastructure.

Virtual Training

Improving virtual training environments by studying embodied sensory/motor perception of event segmentation.

Intentional Systems

Modeling the formation of intentional behavior using autonomous robots, smart environments and eye tracking studies.

SODAS

Self-organized Development of Adaptive Systems

Discrete K-Set Toolbox

A discrete, difference equation implementation of Walter J. Freeman’s K-set neurodynamical models.

Evolving Neurocontrollers

Evolving neural models to control autonomous robots.

Immersive Head Tracking using Wii Controller

Using a Wii controller to provide affordable immersive experiences for virtual training simulations.

Vex / CMUCam Integration

Vex PIC microcontroller integration with CMUCam camera for active vision.

NISL Rocks Cluster Configuration

Moving and configuring the NISL lab cluster for Rocks 5.0

Complex Brain Networks

Small world and scale free properties of large-scale networks and the organization of structure and function in brain networks.

SRNN Event Perception Simulations

Simulation of event perception and categorization using simple recurrent neural networks (SRNN).

Globus / Hadoop / Grid Toolkits

Grid toolkits for HPC and utility/cloud computing infrastructures.