Given the growing necessity of indoor quinoa agriculture and the limitations of GPS in such environments, this article proposes a novel design for a greenhouse seeding robot that employs Visible Light Positioning (VLP) for localization. The robot integrates a seed distribution mechanism, a tillage implement, and a suspension system. To validate the design, the suspension behavior was simulated, structural analysis was performed, and a VLP experiment was conducted by projecting a colored grid with a high-intensity LED through birefringent material. The results showed a stabilization time of 2.06 s for the suspension system. In addition, the robot’s pose was estimated using four machine learning models, achieving accuracies ranging from 77.8% to 99.8% for grid position detection.