Autonomous Hexapod
Fully Autonomous 6-Legged Hexapod with Lidar and AI
Built by Elias Nicolas
Traditional methods of transportation often struggle in unstructured environments, but hexapods provide a versatility that can never be emulated. By coordinating six legs with eighteen degrees of freedom (one in the xy plane and two in the xz plane) we plan to build a fully autonomous robot, designed to solve the accessibility gap in data acquisition. While wheeled robots similar to the ones deployed by Waymo and Grubhub require smooth paths, our hexapod utilizes isolated points of contact to traverse rigid terrain. This makes it an essential tool for advancing data acquisition in disaster zones or unmapped industrial sites where traditional mobility fails. The hardware is centered around a Raspberry Pi 3 (swappable with an Arduino Uno Q), which serves as the brain for advanced AI and sensor integration. To fully utilize its physical capabilities, we will incorporate a Lidar module, a camera, and ultrasonic sensors on the robot’s blind sides to ensure 360 degree obstacle avoidance. To manage the complexity of eighteen servos, we will use two PCA9685 drivers to offload PWM timing and simplify cable management. The main source of power will be a LiPo battery and multiple buck converters to provide stable 6V logic for the motors and 5V for the microcontroller, thus preventing system brownouts during peak movement. For this project we hope to experiment and implement multiple methods of control as we continuously scale up to reach a point where the hexapod is fully autonomous. First, we plan to implement an IoT live streaming website, developed with agentic coding tools such as Claude Code and Codex, to provide a real-time Lidar map and camera feed. Finally, we hope to achieve full autonomy by training the robot in Gazebo, taking inspiration from Waymo’s sensor fusion. Our final product will not just move but provide a level platform for high precision sensors in challenging terrains, ensuring that no environment remains out of reach for autonomous data collection.