This study aimed to design, develop, and validate the RSiSTEM framework, a robotics-based simulation learning model intended to foster students' problem-solving and systems thinking competencies within STEM education. The research followed a two-phase developmental design. In Phase 1, the framework was constructed through a systematic synthesis of literature in STEM pedagogy, educational robotics, and simulation-based learning. Phase 2 focused on expert validation using a 5-point Likert scale. A purposive sample of seven experts with backgrounds in educational technology, instructional design, STEM education, and creativity and innovation in higher education participated in the evaluation. The assessment encompassed five domains: conceptual principles, alignment with problem-solving objectives, alignment with systems thinking objectives, feasibility of implementation, and suitability within the STEM education context. Descriptive statistics, including means and standard deviations, were employed to interpret expert judgments. Results indicated consistently high levels of appropriateness across all domains, with overall mean scores ranging from 4.66 to 4.71. The highest rating was observed for feasibility (x[bar] = 4.71, SD = 0.42), while adaptability to diverse learner levels (x[bar] = 4.29, SD = 0.70) reflected some variability among reviewers. The findings confirm that the RSiSTEM framework is conceptually robust, pedagogically sound, and practically feasible for classroom application. By integrating robotics simulation into structured STEM instruction, the framework offers a validated approach to cultivating higher-order cognitive skills essential for addressing complex challenges in the 21st century.