<p>Accurate position estimation is crucial in cable-driven façade robots. Because the position of a robot is sensitive to rope angle errors, a robust angle sensing module is required. In this study, the sources of the errors in a rope angle sensor module developed in a previous study were analyzed, and a new mechanism was proposed to improve the robustnesss of angle measurements. The design parameters (link length, spring stiffness, and module mass) of the new rope angle sensor module were optimized using the Taguchi method, and the rope angle errors were compensated with an estimation model developed using regression analysis. The design parameters were optimized through experiments, and additional validation tests were conducted to confirm the results. The design criteria needed to ensure the contact between the rope and module when designing an encoder-based angle sensor were determined. A robot application experiment demonstrated improved performance of the optimized module compared to the previous module.</p>