Evolutionary Processes

Evolutionary algorithms provide solutions to a number of problems in the design and development of robots.  We are currently funded by the National Science Foundation to use digitally simulated and physically embodied robots to study the evolution of evolvability.  Working with Professor Josh Bongard at the University of Vermont, we explore how the capacity of a population of robots to evolve is linked to the evolution of the genetic systems that code for complexity in the development and function of autonomous robots.  Title of the grant: "INSPIRE Track 1: Is Evolvability Driven By Emergent Modularity? Biomimetic robots, gene inspired information structures, and the evolvability of intelligent agents".

Members of the evolvability team include the following faculty members at Vassar:  Ken Livingston (principle investigator), Jodi Schwarz, Marc Smith, Nick Livingston, and John Long.  Four students serve as Robot Fellows:  Evan Altiero, John Loree, Jessica Ng, and Joshua Ridley.  We also receive invaluable assistance from members of Vassar's Science Technology Support Department, Carl Bertsche and Larry Doe.