A diversity of animals coordinate an array of appendages with a rudimentary nervous system. It remains unclear how these animals respond to sensory cues and navigate their environment. This feat is particularly impressive in sea stars, which coordinate the action of hundreds of tube feet without a brain. My dissertation focused on the control of sea star locomotion with respect to external stimuli. The first chapter of my dissertation focused on the mechanics of tube feet as an array of actuators. The latter parts of my dissertation investigated the extent to which coordination with respect to external stimuli (submerged weight and light) emerges through the collective forces generated by the tube feet and is influenced by motor commands issued by the central nervous system.
The first chapter of my dissertation focused on the variable mechanics of tube feet as an array of actuators. We used histology, kinematics, and mathematical modeling to find that the tube feet of juvenile Leptasterias sp. have a gradient of variable gearing along the arm of the animal. The distal tube feet that are typically used for sensing the environment have higher displacement advantage than more proximal tube feet. The proximal tube feet that are primary force generators have higher mechanical advantage than distal tube feet. This demonstrates the mechanical variability of an array of appendages used for locomotion. The latter parts of my dissertation focused on how these appendages are controlled as sea stars respond to sensory stimuli.
My second dissertation chapter focused on how the tube feet of Protoreaster nodosus respond to variable loading. Previous studies have shown that tube feet can respond to mechanical stimuli even when excised from the body.
However, it remains unclear whether the mechanics of the tube feet is sufficient for coordination during locomotion.
We addressed this by investigating the effect of weight on the coordination in the tube feet of sea stars and the actuators on our robot and model. We found that more tube feet are gathered in each powerstroke with the increase of an animal's submerged weight. We developed a sea star inspired robot with 10 feet powered by servo motors without centralized commands for coordination. The robot successfully bounced with coordinated feet and increased the number of motors per power stroke when weights were added to the body. We saw similar results in our mathematical model without centralized motor commands that treated 250 tube feet as having spring-like mechanics. The results from the model, robot, and animal experiments together suggest that the tube feet are capable of responding to increased loading by recruiting more feet through the collective forces generated by the tube feet.
The third chapter of my dissertation investigated the directional coordination in the phototactic behavior of sea star tube feet. We placed sea stars in an inverted position so the tube feet touch the water's surface. Sea stars continue to perform simulated walking in this position but without mechanical interactions to a rigid substrate. For this reason, we suspect that the coordination of tube feet while in the inverted position is the result of the sea star's nervous system rather than the mechanics of the feet. Our results indicate that the tube feet in the inverted position are more coordinated under directed illumination when compared to darkness. This suggests the role of the central nervous system in the directional navigation with respect to light during sea star locomotion.
Our integration of robotics, mathematical modeling, and experimental manipulation has shown how a system with multiple autonomous agents can increase it's coordination based on mechanics. We have also demonstrated how this system with independently controlled agents can be influenced by a central nervous system while staying autonomous. Our work can also be a basis for understanding how animals with simple nervous control and complex motions move. By identifying how these animals can control complicated motions with simple algorithms, we can potentially apply these same algorithms into future engineered devices.