Vision-Guided Automated Puzzle Solving Robot: System Integration with Embedded Microcontroller, Stepper Motors, and Additive Manufacturing

This paper presents the development and validation of an autonomous Rubik's Cube solving platform that combines embedded microcontroller programming, computer vision, precise open loop motor control, and additive manufactured mechanics. The system acquires live images of a scrambled cube through a fixed camera, processes them with a lightweight vision pipeline, and reconstructs the cube's state in a structured format. An optimized solving algorithm then converts this state into a concise rotation sequence. The resulting motor commands are transmitted to an ARM Cortex-M microcontroller, where embedded firmware schedules coordinated stepper motor trajectories using tailored trapezoidal acceleration profiles to ensure smooth and accurate manipulations despite the absence of feedback sensors. The mechanical assembly, fabricated primarily with consumer-grade 3D printing and reinforced by aluminum extrusion, features a rack-and-pinion $\mathbf{Z}$ stage, opposing translation sleds, and a central rotary end effector that clamps and rotates cube layers. Experiments confirm reliable operation, strong positional repeatability, and consistent problem solving within a constrained time window. Beyond its technical contributions, the platform offers a cost-effective, open source resource for education, enabling instructors and students to explore the complete perception-planning-action pipeline with readily available hardware and rapid prototyping methods.