Current and recent projects
This page was last updated in 2005, and is under construction now. Current research information can be found in my CV.
| LiquidPiston Internal Combustion Engine |
Replacing metal pistons of a conventional combustion engine with liquid water yields a myriad of benefits! Currently building a business from this technology that was primarily developed by my father, Dr. Nikolay Shkolnik
(OK, so this is deviating slightly from my typical neuro / AI theme... What can you do?)
|
LittleDog Robot |
Developing control algorithms that enable this robot to walk (and eventually to run) on rough terrain
 |
Decoding Visual Stimuli from Recordings in Inferior Temporal Cortex |
How is visual stimuli translated into neuronal activation patterns in inferior temporal (IT) cortex. Specifically, given electrical signals from electrodes in IT, is it possible to decipher what objects the animal is seeing at a given moment?
(Current research at the Center for Biological and Computational Learning / Computer Science and Artificial Intelligence Lab, MIT)
 |
MultiElectrode Array art (MEART) |
"A research & development project exploring aspects of creativity and artistry in the age of biological technologies and the future possibilities of creating semi living entaties that might have an emergent behaviour, learn, adapt and are both dependent and independent from its creator and its creator’s intentions."
(Prior research at the Laboratory for Neuroengineering, GA Tech)
 |
Neurally controlled simulated robot: applying cultured neurons to handle an approach / avoidance task in real time, and a framework for studying learning in vitro |
Little is known about how information is encoded in the brain, and even less is known about how computation and useful data manipulation occurs in living neural networks. The goal of this project was to construct a simulated robot to explore data encoding and processing in living neuronal networks. Information was encoded by varying timings between neuronal input stimulations. Encoding information in this way resulted in a non-linear neural response. This response, if interpreted as a computation, can be used to emulate any logic gate, and thus a Universal Turing Machine. This neural response was used to control a simulated robot in real-time to approach an object if it was too far away, or to avoid an object if it was too close. The animat provides a framework for studying living neural networks at the behavioral level. Such an animat may also be useful to study learning in living neural networks, as expressed by changes in the animat's behavior.
(Prior research at the Laboratory for Neuroengineering, GA Tech)