UCSD Vertebrate Movement Lab’s Contrarian Research on Human Decision-Making Influences Autonomous Vehicle Development
Perrone Robotics, Inc (PRI) announced the company is collaborating with Professor Robert Hecht-Nielsen of the University of California, San Diego’s (“UCSD”) Vertebrate Movement Laboratory (“VML”) and its research team on advanced machine learning methods for Autonomous Vehicle perception and control.
The new collaboration project is based on a groundbreaking method for perception and machine learning for autonomous vehicles and will combine Hecht-Nielsen’s work on artificial neural networks (ANN), confabulation theory, and vertebrate movement mathematics with PRI’s applied experience in autonomous vehicles and robots.
“From the earliest days of our MAX platform, we have anticipated and designed in support for machine learning and AI,” explained Paul Perrone, CEO and founder of Perrone Robotics. “But, this collaboration with the UCSD team will completely extend our existing support.
“Dr. Hecht-Nielsen has very novel and powerful ideas that we believe will compel the entire industry to move forward and, when we successfully harness these concepts, users of the Perrone platform will leverage state of the art for machine learning easily and apply them to their existing solutions,” Perrone continued.
The project’s intended outcome is a new framework for PRI’s patented MAX platform that will apply innovative learning techniques to MAX-based applications, specifically in the driverless car space.
Perrone Robotics will have exclusive access to this project and use it to implement highly competent control of driverless vehicles for automobile, truck, and other ground vehicles.
Over the past 10 years, research carried out by the members of Hecht-Nielsen’s UCSD lab has challenged traditional neuroscience explanations for neuronal computations involved in vertebrate movement. Today’s standard human neuroscience claims that neuronal calculations required for making human movements are carried out almost entirely in the brain.
The VML team’s observed data show that almost all of the neuronal calculations occur within sets of neurons within the spinal column. Further, these calculations take on a mathematical form that is entirely different, and completely incompatible with “Deep Learning” approaches that current automotive AI researchers use.
“We’ve been interested in Perrone’s work with autonomy for some time; Paul and his team have proven that the MAX platform can be applied in multiple domains to effectively control robotic vehicles,” said Professor Hecht-Nielsen. “We are very excited to work with the Perrone team to make these good solutions even better by applying the insights and techniques our team has developed and we expect to see enhanced autonomy through improved decision-making, perception, and finer-grained control of a given platform.”
As part of the collaboration, the UCSD VML research team will publish new research that is expected to start a major new trend in the study of machine intelligence. Work on the new platform will continue through 2018 and into 2020.
Perrone Robotics has developed several autonomous vehicles and is in partnership with leading automakers to enable the autonomous car future. With over 13 years of research and development, Perrone Robotics’ MAX platform enables robots of any size the ability to collect input from sensors, fuse that information together in a coherent picture, and then take actions according to the developer’s preference.