A study on automatic navigation of terrestrial insect-machine hybrid systems

• Main Author: Nguyen, Huu Duoc
• Other Authors: Hirotaka Sato
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Mechanical engineering
• Online Access: https://hdl.handle.net/10356/156545
• Institution: Nanyang Technological University

“Terrestrial insect-machine hybrid system” is the idea of controlling a living ambulatory insect’s locomotion using the electrical stimulation of its muscular or sensory nerves. The hybrid system is argued as an alternative to miniature legged robots, as it inherits the small structure and the fascinating locomotion ability of the insect. Its potential applications are those requiring navigation across complex, obstructed terrains, such as post-disaster search and rescue missions. However, numerous issues are hindering this idea. This study discusses three of them. First, the study attempts to enrich the locomotion control of a darkling beetle (Zophobas morio). The enrichment would allow the hybrid system to diversify its navigational tactics, hence benefiting its operation on complex terrains. The insect’s reaction under the tactile stimulation of its elytra was examined. The outcome hinted at a sideways walking pattern, which is new to the existing literature. The electrical stimulation of the same organ was then experimented. The result implied a controllable motion, in which the stimulation of an elytron provoked a contralateral sideways walk. In addition, this motion was graded by the stimulation frequency that higher frequencies resulted in higher sideways speeds. Second, the study discusses the use of a feedback program in the hybrid system’s navigation. The program was built to direct the insect along a predefined sine curve. It was integrated with a proportional controller to alter the signal steering the insect. The performance of the program was found to be dependent on the artificial control aspect rather than the insect. This result supports two new perspectives: (1) besides the biological viewpoint, the control of the hybrid system can be executed under engineering approaches, (2) instead of searching for one optimal stimulus suited to different individuals, the stimulation can be dynamically adjusted based on the insect’s reaction. Third, the study demonstrates the automatic navigation of the hybrid system in unknown environments, which is absent in the present literature. A simple program, operating without the knowledge of surrounding obstructions, was developed to direct a hissing cockroach (Gromphadorhina portentosa). The program successfully functioned under the presence of simple obstacles owing to the insect’s obstacle-negotiation ability. This ability, however, conflicted with the program’s demand when the obstruction complexity increased. Thus, a new program was established to foresee and prevent the conflict. This program provided harmonic cooperation between the insect instincts and the artificial control. It was successfully exhibited in an unknown arena filled with various obstacles. The study expects its results would bring the terrestrial insect-machine hybrid systems one step closer to practical uses.