I grew up in the second most overpopulated city in the world, Mexico City (22x10^6 inhabitants), where I received my BS in mechanical and electrical engineering from Universidad Iberoamericana. Having been exposed to programming at a very early age and growing up with LEGO and the like, I developed a profound interest in building things, especially in robotics and its associated fields.
However, after college, temptation was too strong and I cofounded an Internet startup, which became very well positioned and enjoyed relative success for a few years. My distraction with this endeavor lasted for a couple of years until intellectual curiosity overtook my financial ambition and returned me to the path of science and engineering.
Funded by the tail of the Internet Bubble, I enjoyed about a year of independent research wondering through the fields of robotics and A.I., eventually focusing on nontraditional neural networks. I began exploring sensory-motor learning based on activity-dependent growth of neural connections with the long-term ambition of embedding the computation in application-specific hardware. This work landed me a teaching position at my alma mater where one thing led to another and I founded their robotics lab. This experience in turn served as a launch pad to Dr. Boahen's Lab at the University of Pennsylvania (now at Stanford) where I am learning from the masters of neuromorphic engineering and exploiting my deep admiration of biological systems while working towards my PhD.
My interest is to populate the world with self-sufficient artificial agents that 'live' and work among us. However, given the present state of technology this will probably not occur during my lifetime and thus, for the time being, I have chosen to focus on what I consider a major obstacle, movement. My research is an attempt to advance the real-world interaction capabilities of robotic systems by focusing on motor control.
My approach consists of an integrative analysis of biological motor control that takes into account all stages of the system. That is, the fundamental roles and functional principles of all components involved in the production of movement, form the musculoskeletal system through the motor cortex.
In particular, my Ph.D. project focuses on the cerebellum, the "little brain", a brain structure essential for motor control that contains half of the cells in the nervous system and that is evolutionary more primitive than the cerebral cortex yet conserved through all vertebrates.
Despite the cerebellum's well-known circuitry and crystalline architecture, its computational role remains elusive. This is partially because of the extremely large number of cells in the cerebellum, the small size of many of its cells and the difficulty of recording from multiple locations simultaneously. My project attacks these problems by building a silicon cerebellum: a large-scale multi-chip system that emulates the physiology and connectivity of the cerebellar cortex, the deep cerebellar nuclei and the inferior olive, all in real time and with access to every cell in the model.
I have completed the design of the chip to be fabricated in TSMC 0.18u technology. I'll get it back from the foundry in September and will then begin testing. In the mean time I'm working on the PCB to program, test and interconnect multiple chips to build the olivo-cerebellar complex. Here is a picture of the chip's layout.
Before working on the cerebellum my focus was on emulating muscle. Click here to read about it.