Gary McHale

About

Gary McHale is a researcher in evolutionary robotics, focusing on how artificial evolution can design sensorimotor control systems for both physical and simulated robots. His major contributions center on the comparative study of evolvable neural networks, particularly GasNets and Continuous Time Recurrent Neural Networks (CTRNNs), applied to bipedal and quadrupedal locomotion. His most cited work, "GasNets and other evolvable neural networks applied to bipedal locomotion" (2004, 30 citations), systematically evaluates these network architectures, demonstrating how GasNets can produce stable, efficient gaits. McHale also advanced the field by incorporating energy expenditure into evolutionary fitness measures (2006, 4 citations), a novel approach that encourages more realistic and sustainable robot behaviors. His comparative studies, including "Quadrupedal locomotion: GasNets, CTRNNs and Hybrid CTRNN/PNNs compared" (2004, 15 citations), provide foundational insights for researchers designing controllers for legged robots. By rigorously testing and comparing neural network types, McHale has helped clarify which architectures are best suited for different locomotion tasks, making his work a valuable resource for students and researchers exploring the intersection of artificial life, neural networks, and robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   GasNets and other evovalble neural networks applied to bipedal locomotion

   30 citations · 2004
2. 2
   Quadrupedal locomotion: GasNets, CTRNNs and Hybrid CTRNN/PNNs compared

   15 citations · 2004
3. 3
   GasNets and Other Evolvable Neural Networks Applied to Bipedal Locomotion

   14 citations · 2004
4. 4
   Incorporating Energy Expenditure into Evolutionary Robotics Fitness Measures.

   4 citations · 2006

Key Collaborators

Contact & Links