Development of an intelligent insect-machine hybrid system and its application to create insect-inspired robot
Insects, such as the Mecynorhina. Torquata beetle, possess complex tarsus structures and claws that allow them to navigate rough terrain. The tarsus, the outermost segment of an insect leg, consists of tarsomeres that resemble hollow cylinders, connected in a ball-and-socket configuration, enabling...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis-Doctor of Philosophy |
Language: | English |
Published: |
Nanyang Technological University
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/170763 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | Insects, such as the Mecynorhina. Torquata beetle, possess complex tarsus structures and claws that allow them to navigate rough terrain. The tarsus, the outermost segment of an insect leg, consists of tarsomeres that resemble hollow cylinders, connected in a ball-and-socket configuration, enabling the tarsus to transition between flexible and rigid states. Our hypothesis is that the flexibility and rigidity of the tarsus play a key role in insects' smooth walking. To test this hypothesis, we conducted two types of destructive tests on the tarsal joint. By removing the socket and cutting the elastic membrane of the tarsal joint, we were able to disable the rigid state of the tarsus, resulting in the claws being unable to grip securely onto the mesh substrate. Conversely, when the flexible condition of the tarsus was eliminated by adding tubing, the beetle struggled to withdraw its claws from the substrate. These results confirm the significance of the tarsal ball-and-socket structure on insect walking.
To further explore the function of the tarsus and its potential use in legged robots, we developed a bio-inspired tarsus that is controlled by a cable-drive system. The bio-inspired tarsus comprises five tarsal segments and two claws. The tarsal segments were designed according to the cross-section of the beetle tarsus, and a novel mechanism of claws was developed based on the primary function of the beetle's claws. This mechanism allows the claws to open and spin inward while the cable is pulled and return to their initial posture when the cable is released. These segments and claws were manufactured using 3D printing methods, and the other parts of the tarsus, such as the membranes and tendons, were replaced with springs and cables, respectively.
The actuator system includes a series of three segments with three actuators, which mimic the coxa, femur, and tibia segments of the beetle forelegs. The inverse kinematics using Denavit-Hartenberg (DH) parameters method was employed to replicate the trajectory of the end of the tibia, which is connected to the tarsus. The bio-inspired tarsus was assembled with this actuator system to create the robotic leg. The demonstration on a mesh substrate showed that the robotic leg not only replicated the movement of the beetle forelegs but also regulated the attachment and detachment of the bio-inspired tarsus from the mesh substrate. Our results suggest that the bio-inspired tarsus could potentially be used to improve the mobility of legged robots as well as create an insect-inspired robot that is capable of traversing uneven terrain.
In addition to the insect-inspired robot, terrestrial insect-machine hybrid systems have been discussed as a potential candidates for operations on complex terrains, such as post-disaster search-and-rescue missions. The algorithms for these systems' autonomous navigation and human detection have been developed, supporting them in carrying out such tasks. The next challenge is to combine these algorithms into an exploration strategy that allows the hybrid systems to seek unknown targets in assigned regions. This study proposes such a strategy, consisting of three phases. In Phase I, the hybrid system would promptly search for the target's information. In Phase II, it would approach the target to enhance the collected information, enabling a reliable target classification in Phase III. The proposed strategy was examined via experiments and simulations. A demonstration was carried out, verifying its feasibility. The demonstrated hybrid system could seek a human inside an arena of 4.8 × 6.6 m2 without knowing the human's position. Limitations and potential improvements were also discussed. This exploration strategy would help to bring terrestrial insect-machine hybrid systems closer to their practical use. |
---|