Max Robotics — Global Service Robot Directory + US Certification

About

Brad Dolin is a computational intelligence researcher whose work sits at the intersection of evolutionary computation, robotics, and distributed systems. His research has focused primarily on developing robust control software for modular robots through the application of genetic programming and co-evolutionary techniques. Dolin's most influential contribution, "Co-evolving an effective fitness sample" (2002), which has garnered 14 citations, investigates innovative approaches to fitness case sampling in evolutionary algorithms, comparing co-evolutionary and random sampling strategies across symbolic regression problems and modular robot control tasks. This work addressed a fundamental challenge in evolutionary computation: how to efficiently evaluate evolving solutions without exhaustive testing. Building on this foundation, Dolin explored generalization in evolved distributed control software, seeking solutions robust enough to handle entirely unseen world configurations rather than overfitting to specific training scenarios. His earlier work on programmable smart membranes demonstrated an ambitious vision for self-reconfigurable robotic systems governed by scalable, genetically programmed distributed controllers. Collectively, Dolin's research advances our understanding of how evolutionary methods can produce reliable, generalizable behavior in complex autonomous systems, making meaningful contributions to the fields of evolutionary robotics and adaptive distributed computing.