A freelance robotics engineer has shared his recent work on Reddit—a Swerve Drive Robot designed for autonomous navigation. The robot features four-wheel independent drive and steering, enabling high maneuverability with three kinematic modes: tank, omni, and Ackerman steering.
At its core, the robot utilizes Brushless DC (BLDC) motors with built-in AB encoders for drive control and DC brushed motors for steering, paired with magnetic absolute encoders mounted above the slip rings. The motors operate in a closed-loop system, with the Velocity feedback for drive motors via ODrive motor controllers and the Position feedback for steering via Spark MAX motor controllers.
The robot’s frame is TIG-welded aluminum, with stainless steel suspension linkages for durability. It is powered by four LiFePO4 batteries (24V for drive motors, 12V for steering). The control system includes Arduino for low-level motion control, Raspberry Pi 5 to interface with the motion controller and sensor payload, and NVIDIA Jetson for AI-driven navigation and sensor fusion. The sensor suite includes the Velodyne VLP-16 LiDAR, OAK-D LR stereo depth camera, and Insta360 camera.
This project presents an interesting approach to autonomous navigation, combining precise motor control with real-time sensor fusion. With its modular design and multi-mode steering capabilities, the Swerve Drive Robot offers a versatile platform for exploring AI-driven mobility solutions. Tech enthusiasts and robotics developers may find its integration of advanced sensors and control systems useful for further experimentation and research in autonomous robotics.
