Biohybrid Navigation through Real-Time Terrain Recognition and Natural Climbing in Cyborg Insect

The repositories for the study "Biohybrid Navigation through Real-Time Terrain Recognition and Natural Climbing in Cyborg Insect" provide a comprehensive framework for autonomous biohybrid research, encompassing a Python-based main navigation program, a specialized sensor dataset, and complete experimental analysis files. The software suite features a (1) GUI for conducting real-time navigation experiments via Bluetooth, implementing both reactive-climbing and terrain-adaptive strategies by commanding antenna and cerci stimulation. This system is supported by a (2) dataset of synchronized 9-axis IMU and ToF sensor measurements, manually labeled across surface, climb, descend, and hole categories, which serves as the foundation for the terrain recognition module's multilayer perceptron (MLP) classifier. Furthermore, the final repository (3) contains processed trajectories, quantitative performance metrics, and visualization scripts, ensuring the reproducibility of all reported figures and statistical analyses while providing a benchmark for future advancements in sensor fusion, adaptive control, and biohybrid robotic locomotion.