UC Berkeley

The interaction between cognitive control and biomechanics has enabled animal locomotion capabilities that far surpass state of the art mobile robot performance. But there has been little research focusing on the interaction of the cognitive processes, like learning and decision making, with the biomechanics of the animal coupled to the mechanical properties of the environment. Understanding the cognitive biomechanical strategies, specifically how cognitive processes may work synergistically with the unique mechanical properties of the organism in its environment, is likely to inspire future generations of bioinspired search-and-rescue and environmental monitoring robots. In this dissertation, two model organisms were used to study the cognitive biomechanics of arboreal locomotion. First, high speed branch locomotion was examined using the American cockroach, Periplaneta americana. This cockroach is one of the fastest runners relative to its body length. This hexapedal animal uses an alternating tripod to run up to 1.5 meters per second, becoming bipedal at the fastest speeds. Second, the Fox squirrel, Sciurus niger was studied. This arboreal specialist exhibits high performance climbing, branch navigation, and gap crossing via targeted leaping. Fox squirrels demonstrated learning to improve targeted leaping performance, a novel parkour-like wall jumping behavior, and decisions on when to leap that was dependent on the mechanical behavior of the branch and geometric properties of the gap to cross. These discoveries give new insights into the cognitive biomechanics of arboreal locomotion, and represent a promising path towards a principled understanding of canopy navigation.