Automated assessment in Industrial Robotics using MATLAB Grader

Abstract

This article analyzes the integration of MATLAB Grader into the "Robot Control and Programming" course at the University of La Rioja, progressing from purely formative use in a laboratory setting to comprehensive course assessment. A formative and evaluative learning path was designed using this tool, consisting of exercises composed of fragmented tasks with increasing difficulty, limited attempts, self-assessment, and weighting. This path was implemented during the 2025-2026 academic year. The quantitative results obtained (22 students assessed) are analyzed, correlating progress during the automated practical exercises with performance observed in traditional final exams, both regular and resit. The data reveal that the technical rigor of the automated assessment acts as a harsh filter, leading to critical dropout rates, especially in advanced modules. This raises a crucial debate about current student persistence and how to approach assessment using scripts in groups of students in courses related to control, automation, and robotics.