Cognitive Networks
Avineon has developed an approach to high-level computing that is a synthesis of fuzzy logic and neural network approaches and was also inspired by some affective computing approaches and work in intelligent control. The cognitive network accepts any number of inputs translates them with fuzzy logic transfer functions and finally mixes them in a neural network-like way to produce a final decision. This decision can be applied to a number of different mechanisms depending on the needs of the application including
 |
 |
 |
 |
 |
Goal-setting for autonomous vehicles |
|
Network intrusion activity monitor |
|
Spam detection |
|
Modeling and simulation of semi-intelligent behavior |
|
Swarm intelligence modeling |
|
 |
 |
 |
 |
The cognitive network approach is still in development. The core approach has been laid out and implemented in several test environments including collaborative control of multiple autonomous vehicles and in the simulation of a simple swarm. Videos of the cognitive network controlled behavior of these simulated systems are shown below.
Avineon continues to actively pursue this technology and will update this site as cognitive network capabilities and applications are expanded. Current areas of research are.
|
|
|
|
|
Improving cognitive network learning and training methodologies |
|
Developing additional autonomous vehicle mission planning and execution capabilities |
|
Pattern detection and classification |
|
|
|
|
|
Company Contact :
Dr. R. Scott Starsman / 757-232-7043 / sstarsman@avineon.com
<<Back |
