This empirical study presents the implementation and validation of a continuous assessment methodology aimed at strengthening autonomous work in engineering students. The intervention was conducted during the first semester of 2025 with a cohort of 70 students enrolled in the "Introduction to Programming" course from the Civil Engineering in Computer Science program at the University of Atacama. The proposed methodology was structured across three levels: strategic (macro), operational (micro), and a third focused on continuous improvement. Implementation was carried out through a didactic intervention cycle that integrated Problem-Based Learning (PBL), flipped classroom strategies, evaluation instruments, educational technology, and the institutional virtual platform. Validation was performed using perception surveys based on a Likert scale, along with performance and participation indicators. Preliminary results show a 25% increase in timely task submissions, an 80% perceived improvement in self-management skills, and an increase in the integral process index, reaching a score of 4.0 compared to 3.28 reported in a previous study. This index incorporates dimensions such as self-management, participation, and task completion. The findings indicate that the methodology constitutes a replicable and adaptable model for other courses that promote autonomous learning. The main limitation identified was the need for more time allocated to individual feedback. Future work aims to extend its application to intermediate-level courses to assess its pedagogical sustainability.